Toby Sauer Kevin WallaceToby Sauer Kevin Wallace, CCIE No. 7945 ciscopress.com Table of Contents ua eet N CL ecand Care a refers Er) ed Cisco Unified Communications Manager Leen ee see ee Sead Pa eacuad Pry Soa Cee ccd eet esean ney rr Coo Croats(21 CNP Voice QVoice 642-437 Quick Reference About the Author ‘Toby Sauer is the lead voice instructor and voice curriculum manager for Skyline Advanced Technology Services. Toby brings 30 years of experience in the traditional voice, data, and VoIP arenas. He has been involved in Cisco VoIP since the beginning, when he was working with traditional VoIP and was involved in the earliest installations of Cisco CallManager “Toby has installed many different implementations of Communications Manager and was responsible for converting most of the Midwests Cisco offices from traditional PBX to CallManager ‘Toby became a Cisco voice instructor in 2000. As the Communications Manager product continued to grow and develop, “Toby was a key instructor to many of the original deployment partners Toby currently holds CCNP-Voice, CCNA-Voice, CCNA-RS, CCSI, and various partner-level certifications. Toby teaches all the Cisco Standard Voice courses and many custom variations ofthese courses. Kevin Wallace, CCIE No. 7945, isa certified Cisco instructor, and he holds multiple Cisco eerifications including CCSP, CCVP, CCNP, and CCDP, in addition to multiple seurity and voice specializations. With Cisco experience dating back to 1989 (beginning with a Cisco AGS+ running Cisco 1OS 7.x), Kevin has been a Network Design Specialist forthe Walt Disney World Resort, a Senior Technical Instructor for SkillSoft Thomson NETz/KnowledgeNet, ane a network manager for Eastern Kentucky University: Kevin holds a bachelor’s of science degree in electrical engineering from the University of Kentucky: Also, Kevin has authored multiple books for Cisco Press. About the Technical Reviewer Alex Hannah, CCIE Voice No, 25883, is certified Cisco instructor, specializing in teaching the Cisco Advanced IP ‘Communications product line. He has more than 7 years of consulting experience in Cisco Unified Communications for SMB through enterprise spaces. He is president of Hannal Technologies LLC, a Richmond, Virginia-based Cisco consulting firm specializ-ing in Cisco advanced IP communications and application development using Microso® technologies, He holds a Bachelor's degree in information systems from Virginia Commonwealth University with a minor in business, Additionally, he isthe founder of UCCX net, a video-based training website forthe Ciseo UC product line. In his spare time, you can find Alex on his boat wakeboarding with his family and friends191 CCNP Voice Goce 642-437 Quick Reference Icons Used in This Book pore Cio ures Prom ‘coeruncatone Wenge ‘Cormanratens errr Bx (22011 Cico Systm Ie lights reserved. The publlction fe protected by copyright Pisce ane page 28 formare detail:U4 CNP Voice QVoice 642-437 Quick Reference Section 1 Introduction to Voice Gateways “Modem enterprise network designs need o support the transmission of voice traffic. While more and more trafic in the business \world is originating from VolP products such as IP phones, ther is still anced to interface with the traditional telephony world. Interfaces such as T1/E1 PRI for public switched telephone network (PSTN) connectivity, analog lines for PSTN backup, and analog stations for traditional devices suchas fax machines and modems ae sil quired, For this purpose, devices known as gateways provide the transition between the packetized IP telephony devices and the traditional tolephony devices, The term Voice over JP, or VoIP, is used to describe the transmission of voice over a network using voice-cnabled routers, The term IP felephony refers tothe use of IP phones and a call-processing server (fr example, Cisco Unified Communications Manager [UCM), However, besause many voie-enabled networks contain both VolP and IP telephony componens, these terms are often ‘sed interchangeably This section introduces you tothe basics of VolP networks Specifically, you willbe introduced to a collection of VolP components and protocols, you review a collection of Cisco VolP router platforms that can act as VoIP gateways, and you investigate approaches for deploying call-outing intelligence across mukiple sites. Understanding Cisco Unified Communications Networks and the Role of Gateways Traditional Telephony The tadtional telephony systems of the past 20 years are evalvin15] Section 1 Introduction to Voice Gateways architecture. While the transition to a fully pcketized form is well under way, there is still anced to interface with the traditional ‘components ofthe circut-swtched environment, Cisco Unified Communications addresses this need witha wide varity of voice gateways ‘The traditional telephony nctwork, which is still widely in use, consists of a nctwork of central ices (CO) that are interconnected ‘through large-capacty voice and data circuits that allow calling most anywhere inthe world today. These high-capacity connections ‘pass information between COs using Signaling System 7 (SS7) ox IP, The COs provide connectivity to single homes using analog ‘phones and larger concentrations of people using Public Branch Exchanges, or PBXs, These PBXs are connected tothe CO using Media Resource > Conference Bridge menu option, as shown in Figure 1-28[86] Section 1 Introduction to Voice Gateways (Click the Add a New Conference Bridge link that appears on-screen. Assume that in this example you are using a router that uses the ‘C5510 chipset. You will therefore specif Cisco 10S Enhanced Conference Bridge in the Conference Bridge Type field, as shown in Figure 1-29, If your router uses the C549 chipset, you instead specify Cisco 10S Conference Bridge. FIGURE 1.29 Speciying the Conference Bridge Type Enler the name you assigned inthe 10S configuration (using the register command) into the Conference Bridge Name field, as shown in Figure 1-30. Select an appropiate device pool, and click Insert.[871 Section 1 Introduction to Voice Gateways FIGURE 1-30 Specitying the Name ofthe DSP Resource ‘conference dridge Configuration au. SSStEni ee Verifying DSPs Several commands area ble inthe router to verify DSP resource configuration and operation. ‘The show voice dsp command is used 1 verify codec complexity configuration. In addition, this command allows verification of DP state such as idle, busy, busyout, usyout pending, bad, shutdown, and download pending, ‘Toccheck the DSP status of DSP profiles, use the show dspfarm dsp ‘command. This command shows DSP status and function such as conf and xcode,[se] CNP Voice QVoice 642-437 Quick Reference Section 2 VoIP Call Legs In this section, you will lea about how VoIP taflc is created. Topics such as codecs and transport layers are examined in more de- tail, Examii ing VoIP Call Legs and VoIP Media Transmission VoIP Overview ‘VoIP wai has both similarities to and differences form tational telephony. The most apparent ilference is the transport method Traditional telephony used circut-switched technology wher the physical wire leading o one device is electronically connected to the physica cirsuit of another device, The technology swvlees (connects) physical cletical irclts, VOIP uses packet switching technology. The source devee creates voice traffic in the form of IP packs. These packets are weated in he same manner as dala packets, Ech router or switch examines the addressing ofthe individual packet and sends it its destination. Whe the packet ‘ives a the destination the end device extracts the voice fom the packs nd plays the voice tothe wae. Both traditional and circuit-switched methods use similar basic signaling functions. Supervisory signaling such as o-hook, address. signaling such as dual-one multiffequency (DTMF) tones, and informational signaling suchas dial tone are very much the same in both environments, Traditional telephony uses a wide aray of signaling protocols. Digital circuits use methods such as ISDN, Signaling System 7 (SS7), and Q Signaling (QS1G), Analog ports use protocols suchas loop-start,ground-start,wink-stat, DTMF, and pulse dialing. VoIP de~ vices use signaling protocols such as H223, Media Gateway Control Protocol (MGCP), and Session Intaton Protocol (SIP) i con- trol call setup and teardown,[89] Section 2 VelP Call Legs ‘Traditional voice uses a dedicated physical circuit to connect devices together. This can bea 4-pair station cable toa phone ora 2-pair ‘TI PRI that can cary 23 calls to the public switched telephone network (PSTN). VlP trafic is carved using User Datagram Protocol (UDP}-based Real-Time Transport Protocol (RTP) packets. There are several RTP flows associated with cach VoIP Components [AoIP network relics ona collestion of specialized hardware and protocols, This section examines some ofthe primary components found in today's VoIP networks. Figure 2-1 shows the basic component of an IP telephony network FIGURE 24 Phone IP Telephony i) Network Gaara UCU Unies ocronreroo ‘na09 Mogeeong Sten Prone ‘Soman phone: Provides IP voice to the desktop, Gatekeeper: Provides Call Admission Control (CAC), bandwidth control and management, and address translation Gateway: Provides translation between VoIP and non-VoIP networks, such as the PSTN. gateway alo provides physical access for local analog and digital voice devices, suchas telephones, fax machines, key sets, and PBXs, (©2011 Cico Systm Ie lights recered. The publloton ie protected by copyright Piso sn page 228 for mare dtl(60) Section 2 VelP Call Legs “Multipoint control unit (MICU): Mixes audio‘vieo streams, thus allowing participants in multiple locations to atid the same conference. Call agent: Provides call control for IP phones, CAC, bandwidth comtrol and management, and address tanslation. A Cisco UCM server often serves asa call agent in a Cisco IP telephony deployment Application server: Pravdes services such as voice mal, unified messaging, and Cisco UCM Attendant Console, Videoconference station: Provides access for end-user participation in videoconferencing. The videoconference station conins a video capture device for video input anda microphone for audio input. The user can view video steams and hear ‘the audio that originates at a remote usr station. Videoconferencing isa fast-growing area of IP communications. On the ozs ca say o » Ta Copan 5 Codecs in H.323 “The H.245 subprotocol is responsible for establishing the media stream between the endpoints. This is accomplished between devices ‘without prior knowledge of each other by means ofa Terminal Capabilities Set (TCS), The TCS is amessage sent from one H.323 de- vice othe other to advertise, among other things, codecs, There is a three-way handshaking on this information. GWI sends its info to GW2, GW2 acknowledges GWI's message and sends its own, and GW1 acknowledges GW2's message Another contribution tothe process by H.245 is determining who wins “disputes” between devices. Amasterslave determination i ‘three-way handshake that sends identifiers suchas terminal type and a status determination number. Terminal types, ordered from highest to lowest, are multipoint contol unt, gatckeeper, gateway, and terminal Ifo devices both use a terminal type of gateway, the highest status determination number wins ‘The final Function of H.24S is exchanging information, such a IP address and RTP port numbers, tobe used to create the media streams. These take the form of fogical chanel commands and acknowledgment. Because the H245 negotiation sends many messages, ths method is sometimes called slow connect. I'he endpoints have prior knowledge ofeach other (customer-owned environment) a method known as fast connec is used. The H.24S information is predster- ‘mined and is sent a par ofthe H.225 setup-and connect messages.[68] Section 2 VelP Call Legs A third method, known as H.323 Early Media, combines the two methods. In Early Media the 245 ie still negotiated, asin Slow Start, but this és done during the setup and connect portions of H.225, H323 Early Media is usualy used with the PSTN, For even larger environments, you can have multiple GKs involved in the cal setup. The main diference with such a configuration is ‘that when the frst GK gets an admission request, i sends a location request (LRQ) and must receive a location confi (LCF) from the remote GK before sending an ACF tothe originating GW. As a network grows in complexity because ofthe number of H.323 end- points, a directory gatskecper can be added. This creates two tiers of gatekeepers. The endpoints register with gatekeepers a the fist level. These gatekeepers have dial plan cnries for their own devices. Al other number requests are set tothe diectory gatekeeper, ‘which forwards the request to the level-one gatekeeper that contols that number range. After the request is forwarded, the destination ceper responds directly tothe requesting gatekeeper “To increase the availabilty of 323 neworks, you can configure muliple GKw/GWS to service the same phone numbers High- svailabilty technologies such as Hot Standby Router Protocol (HSRP) can also be wet maintain uptime in an H.323 network. Configuring H.323 Gateways ‘The following topic covers the commands used to configure routers to support H.323 communication. ‘The H.323 service isan integral par of the VoIP service in the router. This service is enabled by defaul, bat it canbe disabled by en ‘ering shutdown in the global mode voice service voip. Atleast one VolPdiat peer must be esated. Example 2-1 shows a typical dial peer Example 2-1 VolP Dial Peer Roster (contig)# diai-pear voice 2 voip Roster (contg-disl-pocr}# destination-pattern 2468 Rogear(conig-dial-pecr}# session target spr4s19226 “The VolP dial poer uses H.323 by default, H.323 signaling messages are transported over TCP by default and use the destination ad- res listed inthe session target portion of the dial per, Several other aspects forthe H373 functionality ofthe gateway can be modified, Items such asthe session transport, source IP a= ‘ress, and H.323 mers can be changed to suit your environment[69] Section 2 VelP Call Legs “The default protocol for session transport is TCP. Tis canbe changed inthe woice service voip section of programming. Changing ‘this parameter to UDP might be required to communicate with third-party H.323 devices tis very desirable to use a consistent source IP address, especially when programming a gateway to use H.323 with Cisco Unified ‘Communications Manager. Iis best practice 1 use a loopback port for the source IP address for VoIP. Using the h323-gateway voip bind sreacdrs xx.xx.xx.% command guarantees that all signaling and media trafic related to H.323 will use a consistent source IP address, Failure to use ths command can result in one-way voice media traffic, Finally, there ae numerous timers that can be adjusted to suit your aplication. These timers are configured using a voice class h323 ‘command. This creates a list that is applied to the VoIP dial peers. The following lst shows the timers related to H.323 1 voice-class 323 tag: The tag isa unique number assigned to each statement 1% 22S timeout tcp establish: When using H323 gateways with redundant Communications Managers, its recornmended that this value be set 0 3, 22S timeout setup: The default of 15 seconds works w Example 2-2 shows some of the common VolP commands. Example 2.2 H.323 Gateway Tuning Example 223 taeout tep establish 3(01 Section 2 VelP Call Legs ‘These reference shoots cover H.323 GW and GK configuration in Section 5, “Gatckeeper and Cisco Unified Border Element Implementation.” However, for now, you can verify and troubleshoot an H.323 configuration with various show commands, as fol lows: © show gateway: Displays the overall status ofthe gateway © show gatekeeper calls: Displays the curent phone calls thatthe GK panicipated in = show call active voice [brief] Displays the details for current voice calls = show call history voice (last n | record | brief] Displays call evord logs SIP Theory and Configuration SP uses the concep of inviting participants nto sessions, and thos sessions canbe advertised bythe Session Amouncement Projocol (SAP) Like H.3, SIPs a peert-peer protocol. Thee pcr ae caled wer agents (UA), The two types of UA are a fa tows 1 User agent clients (UAC Inte the conection by sending an INVITE message User agent servers (UA: ply to INVITE messages SIP can optionally leverage the Following types of SIP serves: Proxy server: Performs forwarding fora UAC Registrar server: Registers the location of current clients 1 Redireet server: Informs the UA of the next server to contact © Location server: Performs address resolution for SIP prony and redirect servers S1P uses cleartext for sending messages. The two types of SIP messages areas follows = Reques from a client toa server from a server toa client Response: A messFIGURE 2.5 SIF Cal Setup un Section 2 VelP Call Legs ‘A request includes messages such as an INVITE (which request a panicipant to join the session) ora BYE (which disconnects the current call), Conversely, a response message uses HTML status messages. For example, you have probably attempted o connect toa ‘website and received a 404 eror or a SOD error. Those same types of responses are used i the SIP environment Fora SIP clint to got the IP address ofa SIP server, it has to resolve the SIP server's address. These addresses are actually URLs that begin with sip: rather than hp, whichis commonly used in web browsers, SIP addresses can include a variety of information, sucit as usemame, password, hostname, IP address, and phone number. Following isan example of a SIP address In this example, the wser=phone argument specifies thatthe user portion ofthe URL (thats, 18595551212) isa phone number and rot awser ID. SIP devices can dynamically make their addresses known by registering with a SIP registrar server. When SIP devices have their ad- ress registred, a SIP clint might hae the capability to resolvea SIP address by set, perhaps through Domain Name System (DNS) or through a local hos table. However, a SIP client can altematively usea SIP proxy server to query a SIP location database 9 resolve the SIP IP address. Consider abasic SIP call, where one SIP GW communicates directly with another SIP GW, without the use ‘of proxy or redirect servers, as illustrated in. a FroPsiean121 Section 2 VelP Call Legs ‘The basic call setup begins whem a SIP client sends an INVITE toa SIP server (notin that a SIP IP phone can at as ether a server, depending on whether it is originating or terminating thecal) The destination server (hat is, a UAS) responds i itis willing to join the session to which it hs been invited. The originating client (dha is, a UAC) sends an acknowledgment (an ACK message} to the destination server, and at this pont, the RTP streams can flow directly between the SIP GWs (or SIP IP phones). Ifyou introduce a SIP proxy server into your topology, the call setup procedure is similar to tht jut discussed. However, the INVITE is sent othe proxy server rather than tothe destination UAS, The proxy server can consult location server to learn the IP address ‘ofthe final endpoint The destination exchanges call paramiters with the proxy server, which responds to the originating UAC, The UUAC then forwards an ACK through the proxy server tothe destination UAS, afer which the RTP strcam is established ‘When a rediect server is used, the originating UAC sonds an INVITE message to a redirect server, which can consult location server to determine the path to the destination, The registrar server responds tothe UAC witha maved message, telling the UAC the 1P address of the destination UAS. This operation is much like when you conect to a website and you recive a message saying that the page you are looking for has moved toa new URL. You are then automatically redirected tothe new URL. When the UAC leamas the location ofthe destination UAS, a direct connection can be setup between the UAC and UAS. Therefor, the main purpose of a rediect server is to offload the IP resolution procedure from the UAC, one of your SIP servers goes down, the voice network could be rendered unavailable. One way to provide redundancy is o have ‘multiple instances of proxy and redirect servers. Therefore, the UAs can have multiple entes, and ifthe fist server fils, the second server takes over Codec in SIP SIP ike H.323, is considered an umbrella protocol, This means that t leverages other standards-based protocols to provide a wide ‘ange of features. One such protocol is Session Description Protocol (SDP). SDPis an IETF-based format for describing steaming ‘media initialization parameters in an ASCII sting. SDP is similar in function to H.24S inthe H.323 protocol, SDP is also used in MGCP, discussed lntrin this section. SIP leverages SDP to negotiate the type of media audio and video), the transport protocol (RTP or UDP ports), andthe Format ofthe ‘media (audio and video codecs). SIP uses the OfferAnsner model for establishing SIP sessions. Am offer is contained inthe SDP fields tht are sont in the body ofa SIP message, The offer dfines the media characteristics that are supported bythe device (media streams, codecs, directional attributes, IP adres, and ports to use), The device receiving the offer sends an answer inthe SDP fils ofits SIP response with is corresponding matching media streams and codec, whether accepted or no, andthe IP address and port on ‘which t wants io receive the media teams,1731 Section 2 VelP Call Legs ‘There are two ways to exchange the SDP offer and answer messages. These arc referred to as Delayed Offer and Barly Ofer Ina Delayed Offer, the session initiator doesnot send its capabilites in he initial INVITE but wats forthe called device to send its apatilics rst (fr example, the list of codecs that are supported by the called device, thus allowing the calling device to choose the ‘code to be used forthe sesion). This SDP information i sent inthe 200 OK acknowledgment to the orginal INVITE message. The calling device sends its response inthe acknowledgment ofthe 200 OK. The Delayed Gifer method is recommended for SIP trunks because it enables the Internet Telephony Service Provider (ITSP) to provide its capabilitis fs. Early Offer is the default method of SDP exchange in Cisco 10S gateways. In this method, the calling device sends its SDP informa- tion as part ofthe INVITE message. The response from the called device is delivered in the 200 OK acknowledgments to the calling party INVITE, Not all cartirs use the Early Offer method, sits always recommended that you veri what the carrer expects frm the customer equipment. Another possibilty used with PSTN SIP trunks is called Zarly Media. Early Media negotiates the media steams before the call has boen accepted by the called device, This i used when trying to access a voice-driven Interactive Voice Response (IVR) system, when DIMF tones are required, or when an carly audio signal is desied, In Early Modia with Delayed Offer, the SDP information is sent from the called device using either a 183 Session Progress message ‘ora 180 Ringing message. The 183 option is more common. With citer message, the caling device sends a pre-ACK message, as 2 ‘response to either 180 or 183, withthe SDP informs Configuring Basic SIP Use the following two basic steps to configure SIP on a Cisco router: Step. Enable the UA. Step 2. Configure dial peers (Consider Example 2-3 Example 2-3 Basle SIP Configuration Examp! oxter{conig)# sip-ua Roueer{consg-rip-ta)# authentication ueersana doe password sparky Roweer (consvay Section 2 VelP Call Legs Rogter(conlg-dial-pear)# destination-pattern §..- Rogeer{consy_dial-pecr}# session protocol =ipr2 Rogeer{consg-dial-pecr)# session target aip-server In this xample, you enable the SIP UA with the sip-ua global configuration mode command, I'he SIP service requires your device to rgiser, use the registrar command to indicate the location ofthe server. If authentication is required, credentials can be defined ‘with the authentication command. Then, you specify’ the DNS name of your SIP server with the sip-server command, After the SIP server has been defined, in VoIP dial-per configuration mode, you specify tha you want the dial peer to use SIP Version 2, with the session protocol sipr2 command, Finally, instead of specifying an IP address asthe session target, you use the session target sip- server command, which points your dial pecr tothe SIP server that is defined under SIP UA configuration made, Configuring SIP ISDN Support SIP canbe configured for various ISDN features. The common ones are ISDN caling name display © Blocking aller ID when privacy exists Substituting the ing number forthe display name, ithe display name is unavailable ISDN sends information suchas calling name and number in Generic Transparency Descriptor (GID) format SIP reads this inform: tion and ses ito send the calling name information to the SIP endpoint, The default behaviors that the calling number is forwarded by the gateway but the calling name is no. This can be controlled, a5 shown in Example 2 Example 2-4 SIP CLID Forwarding Example voice sacvice vaip1781 Section 2 VelP Call Legs In this example, the signaling forward unconditional command specifies thatthe gateway will forward the signaling payload (name and number) tothe destination gateway. Ithe outbound calling name is desired on an ISDN circuit, the isdn supp-service name call ing command enables this option. ISDN has a privacy setting that is used to protect caller ID, SIP docs not hide the information, It activates a lag and docs not display iton the end device. The caller ID is il vsib in the SIP messages that arc transmitted. In some cases, the incoming call might not provide a name but does provide a calling number. By default, Cisco 10S gateways omit the Display name field i no name isre- ceived It is posible to configure the gateway so that the calling number is inserted asthe display name, Example 2-5 shows the con- figuration for these options Example 2.5 SIP CLID Display Example voice service voip In this example, a global change has been made to substitute the calling number asthe display name fr all calls. This cam also be applied at the dial peer level if more controlled behavior s desired. Dial per 1 contains a command that will prevent caller ID in- formation marked private fom being transmitted to SIP device 10.10.11, This command can also be applied globally under voice service wip. Configuring SIP sRTP Support SIP offers two methods to secure voice communications: SIP Secure (SIPS) using Transport Layer Security (TLS) Protocol and Secure Real-Time Transport Protocol (sRTP), which includes authentication and encryption, Normally these protocols are used to- ether but, starting in 10S 12.422), itis possible to configure sRTP without using SIPS. In this cas, cll signaling should be pro- ‘ected with another security protocol suchas IPsec. ‘The following two examples demonstrate the configuration for SIPS and sRTP, Example 2-6 shows this configuration ina global lay~ ‘ut, and Example 2-7 shows the configuration applied toa single VolP dial pec.81 Section 2 VelP Call Legs Example 2.6 SIP Security - Global jon target Ipveri0.10.1.1 ‘This ‘onfiguration example apples the SIPS and sRTP commands to all VolP alls: sip This command places the gateway into SIP configuration mode trl sips: This command enables SIPS by generating URLs in SIPS format for VoIP calls, sccurertp: This command enables RTP, sccurertp fallback: This command allows the cll setup to fall back to RTP ifthe other device does not support sRTP.Ihis ‘command is not entered and the other device doesnot support sRTP, th call wil Example 2-7 SIP Security - Dial Peer Sccurertp faltback ‘This onfiguration example applies the SIPS and sRTP commands to only calls to devices with numbers 1000-1999: 15 sip: This command places the dial peer nto SIP configuration mode,v7 Section 2 VelP Call Legs wrt ‘This command enables SIPS by generating URLs in SIPS format for VoIP call sccurertp: This command enables RTP, securertp fallback: This command allows the call setup to fll back to RTP ithe other device docs not support sRTP. If his ‘command i not entered and the other device doesnot support sRTP, the cll wil ai, Customizing SIP Gateways ‘There ate other factors that can be modified. The transport used for SIP signaling canbe changed an several levels, the source adress for SIP can be contol timers can be adjusted, and Early Media requests can be ignored session transport {system |p tls | udp}: This command applis to outbound signaling and canbe applied in global mode ‘oF in an individual dial per. The default for SIP transport is udp, Other options ar tep and tep ts, The system atrbute refers to SIP global mode ‘transport [tep tls | udp}: This command applies to inbound signaling and is appli in SIP UA mode, The defaul for inbound SIP transport is that all thre transport types are acceped ua, tep, and tp ths, on port S060. Bind {control | media | all} source-addressinterface-idlipw4-addressjpw+-acdress|ipv6-address jpv6-adéress|: Example bind all sourve-interface loopback 0 ipv+-address 10.11. This command binds control, media, or both toa speciic source interface and IP address. Separate commands can be entered for contol and media, Disable-early-media 180: This command controls whether the gateway will respond to an Early Media request fom another SIP device. This command is applied in SIP U.A configuration mode. The following timers can be adjusted 1 adapt to conditions ofthe network: Connect: The amount of time (in milliseconds) to wait for 2200 response to ant ACK request. Default is $00/ms. Valid range is 100 msto 1000 ms Disconnect: Time to wat fora 200 response toa BYE request Default is SO0 ms, with a range of 100 ms to 1000 ms. Expires: Time for which an INVITE request is valid, Default i 180000 ms, with range of 60000 ms to 300000 ms. Hold: Time to wait for disconnecting a held call by ending a BYE request. Default is 2880 minutes, witha range of 15 0 2880 minutes ime to wait before retransmiting a notify message, Default is $00 ms, with a range of 100 ms to 1000 msvay Section 2 VelP Call Legs Refer: Time to wait before retransmitting a Refer request, Default is 500 ms, witha range of 100 ms to 1000 ms Register: Time o wait befor tetransmitting a Register request. Defaults 500 ms, with a range of 100 ms to 1000 ms © “Trying: Time to wait fora 100 response to an INVITE request, Defaults $00 ms, with range of 100 ms to 1000 ms In addition tothe show call commands that you used for H323 verification, you can use the show sip-ua statistics command to dis- play thre different sets of statistics (thats, SIP response statistics, SIP total trafic statistics, and retry statistics). The following com ‘mands can also be used to verify and monitor SIP components show sip-ua service: Displays te status of the SIP service show sip-ua status: Displays the stats ofthe SIP UA. 1 show sip-ua register status: Display the status of E16 numbers that a SIP gateway has registered with an external primary SIP registrar show sip-ua timers: Displays SIP UA timers show sip-ua connections: Displays active SIP UA connections eplay’s active SIP UA calls In addition to show commands, there are several debug outputs that can be useful when troubleshooting SIP cal show sip-ua eal: The debug ecsip command is very useful for general SIP debugging. This command has many detailed options for tributes such as ports and network addresses. The variation debug cesip messages is most useful for watching cll setup and teardown. The debug, ‘voip ccapi inoout command shows every interaction with the call contol application programming interface (API) on both the tele~ ‘phony interface andthe VoIP se. The debug voip ecapi protoheaders command displays messages that are sent betwoen the ori ‘ating and terminating gateways MGCP Theory and Configuration Although H.223 and SIP are per to-per protocols, MGCP s more ofa lentserver approach to call conta. Specifically, MGCP al tows GW to pinto a centralized call gen for procesing. Ina Cisco environment, this centralized all agent isthe Cisco Unified ‘Communications Manager (UCM) server, as illustrate in Figure 2-6.1791 Section 2 VelP Call Legs FIGURE 2-6 cua ngent MGCP Call Setup. Ge UO Sonat) cate — > “The benefits of MGCP areas follows Alternate dial tone for VoIP environments: MGCP allows competitive service providers to provide service in a ‘competitor's service area, = Simplited configuration for peers (Centralized dial plan in Cisco United Communications Manager peers: Only one dal pcr is needed for each analog port Digital ports donot require dial (Centralized gateway configuration in Cisco Unified Communications Manager Simple Cisco 108 gateway configuration. (QSIG support with Cisco Uniied Communications Manager, “The physical picces that make up an MGCP network, such as call agents, GW, and endpoints, ar called components, However, the logical pices ofan MGCP network, such a calls and connections, ar called canceps. As with any communication protocol, MGCP ‘nas its own terminology, The following ae the terms you will encounter when working with MGCP. 15 Endpoints: An cadpoit is where you interface betwen the VoIP network and the traditional elephony network, For ‘example, an PXS port that connccts ta telephone is considered an endpoint, Endpoint names look much ike an email address[80] Section 2 VelP Call Legs (Gor example, cicuilD@mgcpawciscopress com), These names are composed of to pars, the locally significant name ‘of the endpoint (before the sign) and the DNS name of the MGCP GW (after the @ sign) The locally significant portion ‘epreseats the port identifier, and the DNS name is often the host name ofthe router. For example, S1/SUU/DS1-1(@HQ1 is a ‘TL physically located in Module 1 (S1), Slot 1 (SUI), Channel I on Port 1 (DS1-1) ona router with a host name of HOI Gateways: GWs are in charge of converting audio between a VolP network anda crcit-switched network. For example, a ‘residential GW supports devices that you typically find in residential environments (for example, POTS telephones). © Call agents: call agent isthe intelligence of an MGCP network and controls the GWs and thir endpoins, An MGCP GW ‘sam report evens tothe call agent, andthe call agen an, for example, tl the endpoint what ype of signaling to send othe phone, Recall that an MGCP concept is a logical piece ofan MGCP network. Consider the following MGCP concepts = c = Event: An event is what an endpoint has been instructed (bythe call ‘notice the event ofan attached POTS device going off-haok, calli formed when two o more endpoints are interconnected. ent) to watch for For example, an FXS port might © Signal: call agent instructs an endpoint o send a specific signal when a certain event occurs. For example, afcr the event ‘of a POTS phone going of-hook, the call agent might instruct the FXS endpoint o send the signal of dial tone to the phone MGCP groups rlevant events and signals into packages. For example, a line package contains evens and signals that arc used on subscriber lies, sch as an of-hook event and adial-tone signal. MGCP GW types are defined by the types of packages they support. For example, a trunk GW supports the generic media, dua-ione multfequency (DTMF), trunk, and RTP packages ‘You donot have to canfigure an MGCP GW with dial peers for every destination phone number. Instead, the GW can send each di- ‘led digit to the call agent, until a match is made. However, that approach can put a processor burden on the call agent. An alteative approach is forthe GW to downoad a digit map, whichis a copy of the dial plan that is contained inthe call agent, When a GW has a tly) This list will only exist if authentication or encryption is active. After the CTL file, the phone requests its own configuration fle (SEP caf xml). This file contains information such as Cisco Unified Communications Manager Express IP addresses and firmware, Ths fle wll only exist ifthe phone has already been configured in Cisco Unified Communications Manager Express. If his file does not exist the phone will tr to download the SIP version ofits file (SIP cn,[101] ‘Section 3 Cisco Untied Communications Manager Express Endpoints Implementation Step6 —Ifneither the SCCP nor SIP file exists, the phone will request XMLDefaultenf xml from the TFTP server. This file wll contain enough information forthe phone to verify the firmware level and gain the IP address of Cisco Unified Communications Manager Express, Step7 Verify phone load: The phone verifies the firmware version required by Cisco Unified Communications “Manager Express against the firmware loaded on the phone, Ifthe versions are different, the phone wil r= quest the proper fimuare files from the TFTP server. Step8 Ifthe firmware checks out, the phone will attempt to egister with the Cisco Unified Communications “Manager Express listed in its configuration file, Step 9 Ifthe phone already exists inthe configuration of Cisco Unified Communications Manager Express, the phone will successfully register and Cisco Unified Communications Manager Express will instruct the phone to set up its display information such as buttons, soft keys, and the date and time. Step 10 Ifthe phone does not exist in Cisco Unified Communications Manager Express configuration and Auto Registration is enabled, Cisco Unified Communications Manager Express will build a configuration for the phone and assign the next available directory number, If Auto Registration is not enabled, the registration al- tempt will be ejected Power over Ethernet “Most Cisco Unified IP Phone models are capable of thre different methods of obisining power. These are (in the order of preference) Pome aver Ethernet (PoE), midspan power injection, and wall power PoE and midspan power use two diferent technologies, The original method is called Cisco Prestandard power. This method sup- plies 48 VDC at up o 7 watts of power on the Ethernet cable pairs (1.2.3.6), uses a proprietary method of determining whether device nceds pomer (Fast Link Pulse [FLP)), and is supported by most Cisco phones, Some ofthe newer models no longer support Prestandard power, “The second method isa standards-based method known as 802sa/Pa. This method provides 48 VDC at up to 15.4 wats using command es. This can be configured a the port level w VLAN Infrastructure (Cisco IP Phones contain an internal three-port Ethemet switch, Port 0 connects othe internal phone circuitry Port 1, located on the back ofthe phone, allows connection to an external PC. Port 2, also on the back of the phone, allows connection tothe IP network. When the phon is first plugged in, the phone and the switch ex- art ofthe CDP content from the switch to the phone isthe voice VLAN that the phone should use to tg all “Three methods of packet marking can be use from the phone tothe Ethernet switch: 18 802.10: Inthe voice VLAN, packets ae tagged with a Layer 2 CoS priory value 1 802.1p: Voice packets ar sent in the access VLAN (0) an are tagged with Layer 2 CoS port value 'Untagged: Packets ae sent in he access VLAN, and no Layer 2 CoS tagging takes place “There are three methods of configuring Cisco Ethernet switches to support voice VLANs, Single-VLAN access prt, mult-VLAN access port, and trunk port canal be used. Using a separate voice VLAN provides enhanced security forthe voice trafic, simplifies Spanning Tre, and makes QoS implementation easier. Example 3-1 shows the switch configuration used witha single VLAN access pot Example 3-4 Single VLAN Access Port Router(conis)# interface Rogter{conss)# auitehport mode Roveer(conss)# auitenpert wolee vlan dotip Rogter{conss)¢ awitehport access vias 261[103] ‘Section 3 Cisco Untied Communications Manager Express Endpoints Implementation In this example, the phone will tag is voice trafic with 802 1p (VLAN ID=0). This will allow a CoS marking ofS forthe voice pack- «ls Tis can be leveraged in the QoS policy. Both the phone and the attached PC will use VLAN 261. This is ot the prefered option fora phone with aPC. In Example 3-2, the recommended configuration fora separate voice and data VLAN is shown, Example 3-2 Multl-VLAN Access Port Roater{conss)¢ noxter(cong)# Roter(conss)¢ Roater(conSs}# In this example, the phone will mark the voice trafic with VLAN 262, and the data tafic will remain untagged The switch port will ‘dd the tag for VLAN 261. In some cass, the Ethemet switch might not suppor the configuration shown In this cas, a trunking com- figuration can be sed, as shown in Example 33, Example 3-3 Trunk Port Roterconss)¢ Roater(conss)¢ Roseer{oonss)¢ RoaterconSs}# Roster (con) Rogeer{conss)¢ awatehport trunk allowed 261 “The results ofthis example are the same as those in Example 3-2. Voices tagged with WLAN 261, and daa is tagged with VLAN 262 at the switchport. Because trunk ports cary all VLAN trafic by default, de switehport trunk allowed 261 command restricts VLANs except 261 from this por. The data trafic is using the native VLAN, which is never blocked, ‘You can use the show interface fa xx switehport command to view the operational mode of the port and asigned VLANs,[104] ‘Section 3 Cisco Untied Communications Manager Express Endpoints Implementation IP Addressing and DHCP (Cisco Unified IP Phones require network IP addresses, The IP addresses assigned tothe phones should be assigned from separate sub- ‘ets for easier manageability and security. In most cases, the following guidelines should be followed when deploying IP adresses: 1 Existing IP address subnets shouldbe used for data devices (PCs, workstations, server). "© DHCP shouldbe used to assign IP adresses to Cisco IP Phones. 18 Separate IP subnets shouldbe used for phones 1 Private address space, suchas the 100.00 network, can be used forthe voice VLAN. DHCP can come from multiple sources such as routers, dedicated servers, or in small deployments, Cisco Untied Communications “Manager. The following information nced tobe pat ofthe scope of DHCP: IP address, subnet mask, default gateway, DNS servers (optional), and the TFTP server's IP address. Example 3-4 shows a sample DHCP configuration. Example 3-4 Sample DHCP Configuration gpticn 280 tp 10-111-0.1[105] ‘Section 3 Cisco Untied Communications Manager Express Endpoints Implementation anterftce Fastithernet0/0.211 In this example, two networks have been created, The 101.000 network is for data, andthe 10,111.0.0 network will be for voice. Note thatthe DHCP scope named Phone hasan additional option 180 ip 10.11,041 command. This will inform the IP telephony de~ vices where to find the TTP server. the DHCP servers not on the same subnet asthe IP phones, the ip helper-address xx.xx.xx.xx command is added tothe router's interface that connects tothe voice VLAN. ‘This command will forward any DHCP requests to the address indicated in de command. An cxample of his command's sage is shawn under the “On remote router" heading in Example 3-4 Network Time Protocol Accurate time is very important in Cisco Unified Communications Manager Express. Time references arc used forthe time onthe hones, call lists, voicemail message arial stamps, dcbug ouput time stamps, and Call Detail Records, While the intemal clock of | ‘the router canbe used, this clock is not very accurate and it will rift over ime. Fortis reason, i is recommended that you synchro nize your router toa “master” clock using Network Time Protocol (NTP). Example 3-5 shows a sample configuration for NTP on a Cisco Unified Communications Manager Express Example 35 Sample NTP Configuration clock sanner-tine 1ohe PSP recurring fet sbnaay march 02100 Iast sandy october ©D00 In this sample configuration, the Cisco Unified Communications Manager Expres is configured for Pacific standard time (PST) with an offset from UTC of -§ hours. Daylight saving time is sto automatically change on the dates specified, The server will synchro- nize its clock to two NTP servers, with 10.111 being preferred over 10.222[108] ‘Section 3 Cisco Untied Communications Manager Express Endpoints Implementation Endpoint Firmware and Configuration During the discussion ofthe IP phone startup process, it was learned that the phone downloads a configuration file from the TFT server, which in the case of Cisco Unified Communications Manager Express isthe same router. This configuration fle contains in- formation suchas the IP address of Cisco Unified Communications Manager Express, some phone features such as speakerphone ‘operation, and a firmware filename, The filename, which is sometimes called a laa 1D, is compared to the firmware currently running in the phone. Ifthe frmwvare needs tobe changed, the phone downloads the Toads bin ile from the TTP server. This ile contain the ‘names ofthe actual fimivare files needed fo the upgrade, The phone downloads coc fle and upgrades the phone. The phone resets and goes through the startup process a second time. Several steps must be taken fo make the router ready to transmit the files tothe phones. The firmware package must be downloaded ‘and extracted to the lash memory of Cisco Unified Communications Manager Express. After the files are readin fash, the router ‘must be configured to make the files available using the TFTP server function ofthe router, Use the Wfip-server command to indicate ‘which les willbe available through TTP. The flames are case sensitive. They must match the case ofthe files in flash or the “TTP downloads will fail Example 3-6 shows a sample configuration preparing a router to serve files for 7945 andl 7965 Cisco IP phones uning SIP. Example 3.6 Sample TFTP Server Configuration ‘After uploading the files to flash and adding the tftp-server statements, the lst scp is to add the load command into the Cisco Unified Communications Manager Express programming section, Example 3-7 shows the conf supporting SCCP devices. ration required to make Cisco Unificd Communications Manager Express function asa phone system[107] ‘Section 3 Cisco Untied Communications Manager Express Endpoints Implementation Example 3-7 Sample Cisco Unified Communications Manager Express Configuration for SCCP. telaphony-sarvice “The fll ing list describes this configuration: = telephony-service: this command places the programmer into CME configuration mode 4 coddee G722-64k: This command indicates thatthe default codec for IP phones should be G.722, Because not all phones support G.722, these phones will use the default of G.711 law 1 protocol mode duabstack preference ip: This command configures SCCP phones as t how they will respond to replies fiom DNS. In this example, the phone will look al both IP\4 and IPvé responses but will use the IPv4 responses first. If none fof the IPvt addresses work, the phone will try the IPv6 addresses = ip source-addlress 192,168.01 port 2000: This command configures the IP address and port number that Cisco Unified ‘Communications Manager Express will use to communicate with SCCP devices = user-locale O US: This command sts the default language that shows up on U.S. English phones. If only one language is Preferences, you can selec the network interface MAC address that will be used to identify the IP Communicator You ‘an also choose your own device name by clicking the radio bution next to Use this Device Name and entering any MAC address. "The device name ma start with SEP, for example, SEPOODO0000001, TTFTP forthe IP Communicator can come from the DHCP server, but this will not work ina remote location suchas a hotel room, In this cas, cick the Use these TFTP servers radio button and enter the IP address of the Cisco Unified Communications Manager Express. ‘On the Audio ta, th radio button o use G.725r8instad ofthe default G71 Tula. Managing Cisco Unified Communications Manager Express Endpoints. ‘When one or more phones that are associated with a Cisco Unified Communications Manager Express router are reconfigured, they ‘must be rebooted forthe changes to take effec. Two commands reboot the phones: +The reset command is a hard reboot,ray ‘Section 3 Cisco Untied Communications Manager Express Endpoints Implementation + Must be used for firmware changes, user locale, network locale, and URL parameter changes. + Time-consuming, + DHCPand TETP funetions are refreshed, = Restart: Restart isa soft reboot, Suitable for phone button changes, phone line changes, speed dial changes. + Less time-consuming, + DHCPand TTP are not refteshed. Reset or restart can be applied atthe telephony-service / voice register global configuration modes or atthe ephone / voice register ‘pool portion ofthe configuration “The global reset / restart can be done over a time interval in seconds. This might be used to rest one phone every ten seconds, Verifying Cisco Unified Communications Manager Express Endpoints ‘The following areas can be checked to verify that the ephoncs are geting their proper configuration, They are presented in the order that they are applied tothe phone: = Verify the VLAN ID: The endpoint uses Cisco Discovery Protocol (CDP) to obain the voice VLAN from the attached switch, Use the Settings button on the Cisco Unifid IP Phone to check the VLAN configuration. Verify the IP addressing: DHCP typically provides the IP parameters. Use the Settings button on the Cisco Unitied IP Phone to check the IP-relaed settings, © Verify the files in ash memory: Check and verify thatthe corect firmware files are inthe lash memory of the Cisco Unifed Communications Manager Express router. Also verify thatthe ttp-server commands point to the correct files and are using the same case asthe files in flash. The debug tftp events command will allow monitoring of TFTP downloads. = Verify the firmmare installation onthe phones: Use the Setings button on the phone to check the firme that the phone uses. The debug ephone register command on the Cisco Untied Communications Manager Express also displays which firmware is being installed[122] ‘Section 3 Cisco Untied Communications Manager Express Endpoints Implementation "© Verify the pluone status: Check the phone status using the Stings buton onthe phone, view th dnectory numbers that are assigned tothe buttons, and verify tha the locale information is corect on the endpoint. Verify the sucessful phone registration on the Cisco Unified Communications Manager Express ‘The show ephone and show voice register all commands canbe used to verify phone registration. The debug. ephone register and debug voice register events commands ae useful for watching the rgistratson process as it happens, ‘Because ephone-dns and voice register ds are part of the dial plan, they can be viewed by using the show dial-peer voice summary command, The SCCP DNs will show up inthe 20000 range as POTS dial pscrs, and SIP DNs will display inthe 40000 ‘ange as VoIP dial peers, (©2011 Cico Systm Ie lights recered. The publloton ie protected by copyright Piso sn page 228 for mare dtl[23] CNP Voice QVoice 642-437 Quick Reference Section 4 Dial Plan Implementation ‘Adal plan determines how calls are routed through a VoIP network. In this section, you will see what a dial plan contains and how to create a dial plan that points outward tothe public switched telephone network (PSTN). Having multiple sites can add complexity to adial plan (for example, because of overlapping number ranges). This section also covers potential solutions for such design chal- Jenges. Introducing Call Routing Anumbering plan is a component ofa dial pan. Specifically, one of the functions of a dial pan described carter was the addressing of endpoints, The numbering plan describes the strategy used to asign those adresses, Numbering plans should be scalable, logical and comprehensive, The following sections explore various categories of numbering plans and design strategies for implementing numbering plans. Also, critical aspect of many numbering plan isthe capably to support cmenseny Services Therefore, these sections conclude with discussion of 911 services. Introduction to Numbering Plans Jas s an EP network benefits from a hcrarchical IP addressing scheme, a VolP nctwork can benefit fom a hicrarchcal numbering plan. In addtional to being hierarchical, a good numbering plan should be able to sale for future growth Types of Numbering Plans ‘Numbering plans canbe categori< into one of wo broad numbering (ype: private numbering plans and PSTN numbering plans[2a] Section 4 Dial Pan implementation Private Numbering Plans A prvate numbering plan s used within an organization, so these plans do not have to conform to any extemal standards, Following, area few design considerations for private numbering plans [Number of endpoints: The numbering plan should beable to address all existing endpoints, plus any anticipated growth, ‘while using the fewest number of digits as possible, [Number of sites: The length of site codes should be minimized, while sill accommodating the curent numberof sites, plus any anticipated growth, Direct Inward Dial: Ia block of aumbers is purchased from the local telephone company, the last few digits of that block ‘of numbers might beable to map diretly to intemal directory numbers. This type of arrangement is known as Direct Inward Dial (DID), Ifthe block of numbers purchased is not large enough to accommodate all intemal directory numbers (DN), ‘an auto-attendant might be used to intercept incoming calls and then dicot those incoming ci destinations, 10 appropiate internal ‘Access codes: Whon assigning intemal DNS, ty to avoid beginning the DN numbering with numbers that arc to be used for acces codes, For example, if in an environment, site codes begin with an 8and PSTN calls begin with a9, internal DNs should not begin with an 8 oF 9 PSTN Numbering Plans STN numbering plans can vary widely based onthe county being supported, However, to understand a country’s numbering plan, ‘the inlerational E.164 numbering plan must be considered in adltion tothe country numbering plan. The following comtass these ‘wo diferent categories of PSTN numbering plans: 164: E164 isan ITU standard for an international numbering plan. In this pln, each number has an associated country code (CC), which canbe either one, two, oF thee digits in length. The county code is followed by the national destination code (NDO), whichis then followed by the subscriber number (SN), The maximum numberof digits in an E.164 number is 15. The E.164 name can also be used to refer tothe global dialing plan or what is sometimes referred to as because a+ symbol is added in from of the country code in many parts of the world. [National numbering plan: A country can define its own national numbering plan. For example, the United States uses a ten digit numbering plan, where the fist thrce digits represent the area code, which is formally known asthe numbering plan area (NPA) code, The nest hrce digits ae the office code, also known alternatively as the NXX code or exchange number, and the last four numbers are the subscriber numbers,(125) Section 4 Dial Plan Implementation Following ae common clements fond in national mmbering plans Numbers fr emergency services Directory assistance services Free Is within a geographic region Billing calls to mobile phones (Note thatthe United States is an exception and does not bill cals to motile phones) Suppor for long-distance call, with an accompanying charge ‘Tollfce number ranges Suppor for premium services for example, 900 numbers inthe United States) Support fo international calls Attributes of a Scalable Numbering Plan Scalable telephony networks rquite well designed, hierarchical telephone numbering plans. A hirarchial design has these ive ad vantages Simplified provisioning: Ability to easily add new numbers and modify existing numbers Simplified routing: Keeps local calls local and uses a specialized number ke, suchas an area code, For long distance Summarization: Allows the grouping of numbers in number anges Scala +: Leaves room for future growth 1 Management: Allows conta frm a single management point When designing a numbering plan, thes four atrbutes shouldbe considered to allow smooth implementation 1 Minimal impact on existing systems 1 Minimal impact on users of he system 1 Minimal tansation configuration Consideration of anticipated growth[126] Section 4 Dial Pan implementation Comparing Nonoverlapping and Overlapping Numbering Plans ‘Aal plan canbe designed so that all extensions withina system are reached in a uniform way. That is, a fixed quantity of digits is used to reach a given extension from any on-ne origination point. Uniform dialing is desirable because of is simplicity. A ser does ‘ot have to remember diferent ways to dal number when calling from various on-net locations. Nonoverlapping numbering plan ‘considerations have the following features: Intrasite and intersite calls can use the same numberof digits Require centralized design Impractcal im real life (Careful aumbering design needed fom the beginning ‘Consider an example where acompany has multiple sits that ave been merged into one systom as new companies were acquire: © Headquarters: 1001-1099, 3000-3157, 3368-3985 Branch 1: 1001-1099, 3158-3364 Branch 2: 1001-1099, 3986-3999 While this example may bea litle simplistic, twas common practic in the pas to start all sites at extension 1001 and go up from there. This practice has started to dic out because most people are using DID service. This means tha the ranges of extensions tend 10 be farther apart. “There are thre potential ways to solve these issues "= Redesign the directory number ranges to ensure nonoverlapping, well-structured directory sumbes Use an intersite access code and a site code that will be prepended to the directory number to create unique dalable numbers. For example, you could use an inersite code of 8, ste code of $1 to HQ, a site code of $2 to Branch, and asite code of 83, to Branch2, This would create the number 881-1001 for HO, 882-1001 for Branch, and 883-1001 for Branch2, © Dopot assign DID numbers; instead, publish a single number and use a receptionist or auto-attendant.(a7 Section 4 Dial Plan implementation Comparing IP Routing with Call Routing “The most relevant properties of call routing can be compared to the characteristics of IP packet routing. These similarities are shown in Table 14-1 Table 41. Comparing IP Routing wih Call Routing IP Routing Statieidymamic IP routing table TProute “Hop-by-hop routing (Each router | Inbound and oulbound calllegs. The gateway negotiates VoIP parameters with proved ‘makes an independent decision ) _| and next gateways before a calli forwarded. Called number, matched by destination-pater, is one ofthe selection criteria Most explicit match ale “The ot expt math for destination parm exis, but fer cite ar cond Destination-based routing Preference can be applied to equal dial poor, [Fall entera are the same, random selection, Equal paths I Possible. Often points to extemal gateway or gatekeeper Default route Understanding Dial Plans Dial plans typically organize a group of phone numbers in hicrachial fashion. Consider the North American dialing plan, which consists of ten digits, an example of which fllows: 859-555-1212 “The fist thee digits thats, 859) indicate an arca code, wich is wpically associated with a geographic location within North America, The following thse diits (hat is, 535) are the cena ice (CO) code (thats, the NXX. cos), which identifies a central ‘office location within the area that is specific by the arca code. The final four digits (that is, 1212) point the lal CO to a speifc local loop that goes out oa subscribers physical location, ing sections identify the charetcrisics of a dal plan. Each ofthese charocteristics is then discussed in more det ‘The fll[128] Section 4 Dial Pan implementation Dial Plan Elements A al plan contains elements that perform the following functions © Assigning endpoint addresses: Endpoint addressing determines the format of phone numbers, ‘over an IP WAN (as a prefered path), bt calls = Seleting a path: You might, for example, configure adit plan to place cal might be routed over the PSTN asa backup, 1 Manipulating digits: Dialed digits and the digits making up a caller ID string might need to be manipulated when calling between phones. For example, digits suchas arca code and office code digits might need to be added when calling ou othe PSTN, or one dal string (or example, a 0 10 reach a company operator) might need to be replaced with another dial sting (orexample, the actual directory number ofthe operator's phon). Applying cal restrictions: Cal restrictions can be configured to control which destinations phones are allowed to call, For ‘example, you might not want a lobby phone to beable to call an intemational number = Supporting call coverage: The call coverage featur allows a group of phones, sometimes called a hunt group, to handle incoming calls (for example, calls coming into a cal center). Assigning Endpoint Addresses Endpoint addressing assigns directory numbers (DN) to devices, such as phones. Also, endpoint addressing maps internal extensions to incoming Direct Inward Dial (DID) numbers. However, if you do not have a DID number to map to each internal DN, an auto- attendant canbe used to take an incoming call and route that call to an appropriate internal DN. ‘One ofthe biggest challenges with endpoint addresses occurs when you have multiple sites, and those sites have overlapping DNs, as itustrated in Figure 4-1, Notice that both the Kentucky and Arizona locations have DNs of 1500 and $101FIGURE 41 Overlapping DNs [129] Section 4 Dial Pan implementation ‘Toallow callers in Kentucky and Arizona to have ther alls extended tothe appropriate locations, the administrator could requte the use of site codes, where the dialed numbers prepended with a site code when calling a remote site. In this example, fa caller at DN 1501 in Kentucky wants to call DN 1502 in Arizona, the caller can dial 8201502, Digit manipulation mus then be performed to stip ‘the $20 site code fom the dialed number. so, a good practice isto manipulate the caller ID number to prepend the Kentucky site ode of $1010 1501, Asa resul, the recipient of the call, at DN 1502 in Arizona, would look at his phone and see 8101501 displayed ‘asthe caller ID, This provides the called party with the exact number he should dial to cal back the calling party Selecting a Path ‘The call routing and path selection componcat ofa dial plan dictates where and how calls shouldbe routed through a network, Usually, how a call is routed depends onthe dialed number. For example, was the dialed number on-nt or off-et? ‘An IOS router acting as an H 323 gateway makes these call routing decisions based on dial pocrs. Asa result, in some larger deploy - ‘ments, the numberof dial pers can be large. Recall that multiple dial peers can be configured to point othe same destination, and the preference command canbe issued in dial-peer configuration mode to influence which dial per is used. The prefered dial peer could ‘therefore point across an IP WAN, while alesserpreferred dial peer could point outward tothe PST.[130] Section 4 Dial Pan implementation Manipulating Digits Dial plans also need to accommodate digit manipulation, For example, when a call comes in, the called number nosds io match a des- ‘ination pater thatthe router knows how to reach. Also, you might want the caller ID of the calling number to appear atthe destina- tion phone ina format that will allow the called party to dal back the calling party without having to modify ehe number inthe calling lists ofthe phone For outbound calls, you nce to present a valid dial string to, peshaps, the PSTN, In addition, you might nced to support 911 calls, ‘which might requte Centralized Automatic Message Accounting (CAMA) trunks, A CAMA trunk can help preserve caller ID infor ‘mation being sent out on an analog trunk. For example, suppose a university spans several square mils and has several buildings, If a caller in Building Z calls 911, but the analog trunks coming in from the PSTN are located in Building A, the location information seat ‘out tthe 91 publicsafety answering point (PSAP) will elect the location of Building A rather dhan Building Z. CAMLA can help solve this issue by transmitting the orignal caller ID over an analog trunk to the PSAP, While this is a digit manipulation example, E911 is avery serous issue that must be planned, Besides the use of older technology such as CAMA trunks, digital circuits such as TI PRI and E1 PRI can send a modified CLD. To address the issues unique to IP ‘phones, products suck as Cisco Unified Emergency Responder can track where a phone is lvated inthe IP network and manipulate the CLID ofthe calling device to match its physical location. Applying Call Restrictions You probably do not want your users calling any destination they choose, suchas intemational numbers, 900 numbers (thats, re- ‘mium service numbers), or perhaps even $11 calls (thats, directory assistance calls). Ina PBX environment, the feature that supports Setting call limitations is commonly refered to as class of sevice (CoS). Ina Cisco Unitied Communications Manager (UCM) cnvi- ‘ronment, paritions and calling scarch spaces are commonly used to apply call restrictions. Also, th class of restriction (COR) feature ‘adds call restiction support for 1OS-hased voice gateway. ‘Supporting Call Coverage The call coverage component of a dal plan helps minimize the number of dropped incoming calls. you are not at your desk, for ex- ample, you might have your phone forwarded to another phone Ina small sales envisonment, the goal so intelligently distribute incoming calls across multiple customer service agents, The phones ‘ofthese customer service agents can belong toa hunt group. However, callers donot directly dal one ofthe customer service agents Instead, they dala hunt pilot number, which distributes incoming calls among the hunt group members[31] Section 4 Dial Pan implementation Call coverage can range from a hunt group in Cisco Unified Communications Manager Expres or Cisco Unified Communications “Manager toa full Contact Centr software suite such as Unified Contact Center Express or Unified Contact Center Enterprise PSTN Dial Plans ‘Configuring a dal plan to point ouard to the PSTN can bea complex task Often, you need to use solutions provided by UCM, Cisco UCM Express, oran 10S-based voice gateway. The following sections discuss PSTN dial plan design considerations, and hey examine the synax used to conigure PSTN dia plans PSTN Dial Plan Design Considerations Call routing and path selection shouldbe set up in bot the incoming and outgoing directions. To make the linkage between the PSTN and the internal VoIP network transparent to the end users, and to present appropriate caller ID information to both the called and call- ing panics, digit manipulation might be required. ‘Consider the example shown in Figure 4-2. The topology shows an outbound call coming from DN: 1500 inthe Kentucky office and destined fora phone onthe PSTIN witha phone numberof 480-855-1335,FIGURE 4.2 (Outbound PSTN Call, Example 1 ‘destined fora phone onthe PSTN witha phone numberof 480-555. [132] Section 4 Dial Pan implementation EN marnes be a) Sing and ses a he @ Ba 1 Rau ope 95: DN 1500 in the Kentucky office dials 94805551345 to reach a phone on the PSTN. The leading 9 indicates to the UCM sexver that this cll should be routed out the PSTN gateway: At this point, the Automatic Number [Identification (ANI; that is, caller ID information) is 1500, and the Dialed Number Identification Service (DNIS; that is, the dialed number) is 94805551345, ‘The UCM server matches the DNIS against one ofits route pattems. The UCM server is also configured to strip the leading 9 from the DNIS before sending the call to the PSTN gateway, AC this point, the ANI is 1S00 and the DNIS is 4805551345. ‘The KY_Router prepends the ANI of 1500 withthe local area code and office code. The call is then forwarded ‘out to the PSTN. AC his point, the ANI is 8595851500 and the DNIS is 4805551345. ‘The destination phone (that is, 4805851345) rings, and the caller ID that appears on the destination phone is 8598551500,FIGURE 4-3 PSTN Dal Plan Confguration Example [133] Section 4 Dial Pan implementation ISDN Dial Plan Considerations ‘When using Iterated Services Digital Network (ISDN) trunks, you need to be aware ofa couple of addtional design considerations: © AmISDN network might represent the ANI number asthe shortest information. As a result, you just add onthe PSTN access code to this ANI number, you might not have aval ‘an be called back. You ean, however, use digit manipulation o overcome this issue. Jnble number, along with ype of umber (TON) umber that © AmISDN network, or PBX, might require that when a call setup message i sent to them, rumbering-pln information and ‘TON information needs tobe included. Dgit-manipulation commands can again be used to preseat this type of information to the ISDN network or PBX, Configuring PSTN Dial Plans Figure 43 shows the syntax that might be used when configuring» PSTN dial plan eetyNOTE Sesion, “Gatekeeper ‘nd Cisco Unified Border Element implementation, ‘ge deeper into the ype tsed wo configure ‘oie talon pris. [134] Section 4 Dial Pan implementation This PSTN dial plan has two primary requirements: 1B Prepend outgoing ANI (that is, caller ID) information with the area code and ofice code ofthe Kentucky office © Strip the arca code and office code from the incoming DNS (hati, the dialed number) so that only the four rightmost digits from the DNIS are used to route the call within the Kentucky office, Prepending Digits to the Outgoing ANI Example 41 shows the symtax that satisties the fist requirement, prepending outgoing ANI information with the area code and office code ofthe Kentucky office (hat is, 889555), Example 4-1 Prepending Area Code and Office Code Information to Outgoing ANI y_nouter (config)? voice transiation-rule 1 rx Roster (ctg-tranelation-rale}# rake 2 /*2/ /0SS5582/ Roater(=f9- Router (contig) # Roster (ofg-translation-profile)# exit x Rostar(contg)# votee-port 1/0/0123 Fy Roster (contg-voteaport}@ tranelation-profite outgoing ANT-OUr In Example 41, voice translation rule-1 matches a dal string beginning with a 1 and replaces the I with 859551, So, this rule ‘would replace the amber of 1500 with 8598551500, The voice translation rule i them applied to a voice translation profile named ANL-OUT. Notice the calling parameter in the translate calling 1 command. The calling parameter means that the translation will be performed on the calling number, which isthe ANI num- be, Finally, the voice translation profile of ANI-OUT is applied in the outgoing direction to the T-based PRI, The D channel of an ISDN PRI circuit bil on 1 voice port 1/00 is referenced as voice port 1/00:23, Stripping Area Code and Site Code Information from the Incoming DNIS Example 42 illustrates the commands used to remove the area code and site code information from the DNIS number coming in from the PSTN,[135] Section 4 Dial Pan implementation Example 4.2 Digit Stripping from the Incoming DNIS. Hy Routar (config) voice translation-rule 2 Roster (ofg-translation-rule)# rule 1 /°8595551/ /1/ Roster (afg-tranelation-rule)# endt ‘Roster (conlig)# voice translation-profile DH Ny Router (cig-translation-profile)é translate called 2 vy Roster (ctg translation profile)# exit Ry Rovter(cig-translation-rale)# vaice-port 2/0/0123 x Roster (confg-voscepore)# tranelation-profite tncaning DHT Example 4-2 uses woice translation rul-2 to match a number beginning with 859551. That string of digits is thn translated to the ‘number 1. Therefore, if DNIS information of 8595551500 were coming into the KY _Router gateway, voice translation rule 2 would ‘replace the leading 8598851 digits with aI, resulting in a DNIS of 1500, Voice translation rule 2is thea applied tothe voice translation profile DNIS-IN. Note the called parameter inthe translate called 2 ‘command, implying that the translation rule is applied to dialed digits (hat is, DNIS information). Finally, the voice translation profile is applied in the incoming dicction to the D channel ofthe Tl-based PRI circuit Monitoring and Troubleshooting PSTN Dial Plans Cisco offers a collection of show and debug commands for monitoring and troubleshooting dial plan configuration, as shown in Table 2, Table 42 PSTNDi an Monitoring and Troubleshooting Commands Command [[Deseription show dial-peer voice ta Displays detailed configuration information bout the specified dial peer Displays summary information (fr example, typeof dal peer, destination pater, port, and ‘show dial-peer vole SummarY | session target) forall dil peers configured on a router show dialplan numberdial- | Shows which dial peer would be wed by the router if call were placed to the specified dial string string[138] Section 4 Dial Pan implementation tcbugida ost ‘Dips alin SDN Layer galing segs, which actos ANT DNS debug voip dialpeer Displays real-time dil-peer matching debug voice transtation Displays real-time operation of voice tran debug. voice eeapiinout Displays call control messages test voice transation- rule | Displays the results of passing a umber through a specific translation rule Integrating Private Numbering Plans with PSTN Numbering Plans Following are design considerations for integrating a company’s private numbering lan with a PSTN numbering plan: Variable-ength dial strings: When calling PSTN numbers ftom within an organization, the length ofthe dal strings can vary based on the destination, For example local call might not require an area code, whereas a long-distance call docs Tacrefor, digit manipulation might be required to support PSTN fallback, while making the numbering plan complexity as ‘wansparent as possible to the end users. Centrex support: For some smaller olice environments the local telephone company can provide PBX-like features for the ‘office location, without requiring the office to have a PBX or key system, These Contrex services typically have DNs that are four or five digits long, This can add significant numbering-plan complexity when a call comes into. a VoIP network fom the STN and that calls routed out to the Cenex environment. Support for voicemail: Some voicemail systems use a diferent numbering plan than the telephone system they support Asa result, digit manipulation might be required t support transfers to voicemail PSTN fallback: Ia call cannot be placed over the IP WAN, pethaps because ofa lack of bandwidth or because ofthe IP \WAN being unavailable, that call might be able tous the PSTN asa backup path, However, the orginally dialed digits (for ‘example, aN ora DN prepended by’a site code are typically not sufficient to route the call through the PSTN, Therefore, digit manipulation needs to be performed to prepend appropriate digits, such as an area cod, office code, o an access code Internationa calls: Because country codes vary in length, numbering plans must be constructed to support the international destinations that need to be calle.NOTE Network desma should be aware of and fallow oe, municipal, tate, sd feral laws reg sng 91 sence 1137] Section 4 Dial Pan implementation Even though a VolP network implementation should strive to make the transition as wansparent as possible to end users, some aspects of how end users place thir calls, interpret caller ID, or check ther voicemail might be different. Therefore, a critical implementation step is educating the end users abou the operation ofthe newly deployed VoIP telephony system, [Number normalizatio : Ina large VoIP network, such asa VoIP network supported by service providers, diferent sites in the overall VoIP network might us different DN lengths. To support calling from one site to another ste over the VolP ‘network, number normalization can be used, The ides of number normalization is that when calling between sites, the calling and called numbers arc modified to a standardized format (suchas 10-git dialing). Equipment atthe individual sites can ‘then perform digit manipulation to creat appropriate calling and called dal strings for use within those individual sites. Understanding 911 Services Although understanding the implementation of 911 emergency calling is beyond the scope ofthese reference shes, vou should understand some ofthe basic components that allow callers to have their location information corectly communicated toa 911 ‘operator. Following isa listing of some ofthese basic components Automatic Number Identification (ANI): The ANI isthe phone numberof the person placing thecal Automatic Location Identification (ALI): The ALI is «database that associates a phone number witha physical location, Updates to this database can tke about 48 hours. Therefore, updating the ALI database in a mobile environment is not an effective solution, Publ safety answering point (PSAP): The PSAP i where a 911 call isterminate. Emergency-response location (ERI): The ERL is used in mobile environment to identify the approximate locaton from ‘which an emergency call was place (or example, a specific lor in a building) Emergency Location Identification Service (ELIN): The ELIN is used for mobile environments, Specifically, the ELIN ‘phone number replaces the ANI phone number when an emergency cal i sent to the PSAP. Master Street Address Guide (MSAG): The MSAG isa government-maintained database that maps geographic regions 10 PSAPs, which are responsible for handling calls coming from those regions[138] Section 4 Dial Pan implementation § Selective router: A selective router i telephone switch that routes 911 calls to appropriate PSAPs, based ona calls ANI information, Centralized Automated Message Accounting (CAMA): ACAMA trun is an analog trunk that connects a customer's ‘hone switch directly to a selective router A CAMA trunk carries only 911 calls, and inthe case of am analog trunk, if CAMA ‘were not use, location information visible tothe PSAP might be incorrect Describing Digit Manipulation ‘ACisco Unified Communications Manager (UCM) server fers several features to manipulate digits, select appropriate path, and limit which destinations various phones can call, However, IOS voice gateways might also need to perform these types of features, The following sections, therefore explore the theory and configuration for thes types of 10S features. Digit Collection and Consumption Ian endpoint ends aing digits one by one, Cisco United Communications Manager Express stats digit analysis immediatly upon receiving he it digit ‘By cach additional digit that is received, Cisco Unified Communications Manager Express can reduce the list of potential matches (tat the cal-soutng tbl entries that mach the digits that have ben recived so far), After a single en, suc as directory mum ber 1001, is matched, the so-called cuenta is uscd andthe eas sent he corresponding device Cisco Unified Communications Manager Express does not always receive dialed digits one by one. Skinny Client Control Protocal (SCP) phones always send digit by digit. This behavior can vary abit between models of phones. Some phones such asthe 7965, will use dit by digit when dialing through the dil pad, but it will use en bloc when the redial softkey is used, On an older model such as the 7960, digit by digit is always used. Session Initiation Protocol (SIP) phones can use en bloc dialing o send the entire di= ‘led string at once, or they can use KeyPad Markup Language (KPML) to send digit by digit. I digit are received en bloc, the entire ‘received dial string is checked at once against the dial pan ‘Table 43 points out the differen dialing methods[139] Section 4 Dial Pan implementation Table 43. PSTN Dial Plan Monitoring and Troubleshooting Commands Device ‘Signaling Protocol ‘Addressing Watiod Digitby digit sceP En bloc (Type B phones only) IPhone En bloc sip KPML (Type B phones only) SIP dial rules En bloc Gateway MGCPISIP 323, (Overlay sending and receiving (SIDN PRI only) As evidenced inthe preceding section, digit manipulation i often required in voice gateways to create appropriate caller ID and di- led numer information, Digit manipulation encompasses several features, including prepending digits to a numbr, removing digits from a number, and translating one number to another number. These numbers canbe such things a caller ID or dialed number infor- Using the digit-strip Command ‘When using a plain old telephone service (POTS) dal peer, by defo, only digits matching wildcard characters in the destination- pattern command are forwarded out ofthe POTS pot. For example, consid Figure 4-4 The analog phone is calling a number that ‘that router should forward to an attached PESX. The phone sends the digits of 4123 tothe router. However, the router has a destina- tion-pattera command that matches 4, and because only the digits 123 match wildcards (hats, the periods) the 4 is tipped of the string sea tothe PBX.[140] Section 4 Dial Pan implementation 28 > 128 > ‘he diit-trip command canbe used to override default digit-stripping behavior In this example, the no digit-strip command could have been issued in dial-pocr configuration mode to prevent any digit stripping, Using the forward-digits Command Another approach to overriding the previously described dgi-stripping behavior isto use the forward-igits command. Specifically, the forward-igts command canbe issued in dia-pser configuration mode, fllowed by a number. The number indicates the number ‘of digits that, beginning a the right of the dil string, shouldbe forwarded out of POTS port. In this example, the forward-ligits 4 ‘command could have been used to cause the router's dal poor to Forward all four of the dialed digit Using the prefix Command The prefix command, followed by one or more numbers, could be used to prepend those numbers to a dal string, For example, POTS dil poor witha destination-pttem of 9011 would, by default, strip the 9011 before forwarding the cal tothe PSTN. Using the no digit-strip command would not work, in this case, because it would send the 9 along withthe O11. The forward-digitscom- ‘mand would also not workout hee because the length ofan international number can vary i length. In this case, itis best tallow ‘the default behavior af the POTS dial peer remove the 9011 and add the 011 back on using the prefix O11 command. However, the prefix command could also be used in the previous example, Because the default digi-stipping behavior ofthe dial peer stripped off the digit 4, the prefix 4 command could be used to replace this stripped digit Using the num-exp Command The number expansion (num-exp) command can be usd to replace one number with another number, For example, consider a tele- ‘commuter working from home, as depicted in Figure 4-S. Eventhough the telecommuter could be teached through the PSTN by dis ing 855-1345, you might want this telecommuter’s phone tobe reachable through a four-digit umber, like the other intemal numbers[at Section 4 Dial Pan implementation ‘You could therefore use the num-exp command o replace an internal directory number (DN) with the public switched telephone ‘network (PSTN) number For example, if the telecommuter were assigned a DN of 2020, the global configuration mode command of rnum-exp 2020 8551345 could be issued to take the dialed numberof 2020 and replace it with a dialed number of 851348. Be aware, ‘however, that you stil need a dial peer that can match the newly created dal string of $5S1345, FIGURE 4-5 Tekonmur's The mum-exp row Command Cor The clid Command In Integrated Services Digital Network (ISDDN) circus, there i calling party aumber information clement (IE) that is sent through the Q931 protocol that is used to send caller ID (or CLID) information. This information element can include two diferent calling vided, or unscreened, number. The other one isa network-provided number. I'you want to manipulate this caller D information, you can us the clid command in dal-per configuration mode sumbers. One i a wser Forexample the eid network-number umber command enables you tose the ntwork-povided number inside of Q 931 infomae tion clemenis. You can we the eld second- number strip command to remove the ise provided numb fom th information cl- smcat Als, th elidrestit command ses the presentation it inthe Q93] message to preven! the display of CLID information. ‘You can also remove the numbers completely by using the elid strip command twice, to remove not only the calling number but also the calling name. Specifically, you enter elid strip to remove the calling number You then enter eld strip name to remove the name.[142] Section 4 Dial Pan implementation Example 4-3 shows a sample dial plan with several different types of manipulation. It should be noted that digit manipulation can ‘happen at many levels. This is one example, but wth some thought, there could be several other ways to gain the same resuls tis important to keep in mind the digits th end user will dial and what digits are roquired atthe destination, whether that sa emote router ofthe PSTN. Example 4-3 Digit Manipulation Example pret pore 9/223 dertination-pattars 3. eee 13129253 port o/as2a no digit-serip pore a/a23 is example, we have a branch oflice with extension numbers 4000-4999. This fice is currently on a raditional PBX, but the so that the end users will not have tobe retrained when they move to IP telephony “The mum-exp 4. 408SSS4,.. command changes any 4000 series number into 4085554000, This number has to be matched tothe dial plan a second time, and a match i found with dial peer 4000 pots, Tis dial per is pointed to the local PRI Because the default (inthis ase, 4088554), the programmer used the forward-igis all ‘command to mullfy that behavior. In addition, the digit 1is prefixed to the ten-igit number to gain compliance with the PSTN's dial ing requirements,[143] Section 4 Dial Pan implementation The company has another office with a telephone number range of'3125583000 through 3125853999, Tis location isan IP-based lo- cation with both WAN and PSTN access. ts desired that the WAN is usd frst andthe PSTN second, The end users will dial 3000— 3999 to access tis site, They wil ot have to dal the PSTN number ifthe WAN is unavailable Two dial peers are created The frst willbe a VoIP dial peer, and the second will he POTS. Both dial peers will se the samme 3. pat- tem. To control which dial per is used first the VoIP cal per uses the default preference 0 (hidden by defaul) command, and the POTS dial pec will use preference 1 the ist ial per is matched, the calls routed unchanged tothe remote VoIP target, that call setup fails, the second dial peer will route thecal othe local PRI after compensating forthe default digit-strip command by prefxing 13128883 in from ofthe lst three digits ofthe original number The final dial peer is for emergency calling. Because all hee digits (911) would be stripped by default, the programmer canceled the default behavior by isuing the no digit-strp command, Using Voice Translation Rules and Voice Translation Profiles ‘The most advanced of the 10S-based digit-manipulation approaches involves voice translation rules and voice translation profiles, Specifically, a voice translation rule can define ast of rules (as many as 15) to change digits, ISDN numbering type, and number plan. These voice translation rules are associated with a voice translation profil, which can reference up to three voce translation rules: = Aruleto translate the dialed number = Aruleto translate the caller ID information = Aruleto translate the redirected called number These profiles can be applied o, for example, voice pars or dial peers Each ofthese emits can have two voice translation profiles applied, one forthe inbound direction and one forthe outbound direction ‘A ransation rule uses regular expressions to perform digit matching, Table 4-4 provides a paral lis of some ofthe more ‘commonly used regular expressions.[14a] Section 4 Dial Pan implementation Table 4-4. Commonly Used Regular Expressions Regular Expression ‘Description ‘Matches a number at the beginning ofa string of numbers ) ‘Denotes the beginning and ending of matching and replacement strings ina voice translation rule \ [Negates the special meaning ofthe next character so that the next character will be interpreted a the literal character “Matches one digit ‘Matches the previous character 0 or more times + ‘Matches the previous character I or more times 10 Groups elements of regular expressions into sets ‘To belie understand the use ofthese regular expressions, consider the voice Wanslation ral showa in Example 4 Example 4.4 Voice Translation Rule Example (cons)? voice transiation-raie 1 Rogeer(oeg-tranalation-rule)# rae 2 /MBSS\(2..-\)/ /212\3/ In Example 44, the wo ses of forward slashes indicate the matching string and the replacement string, respectively, as shown in Figure 4-6 (22011 Cico Systm Ic ligt reserved. This publication protcte by copyright Pisce sme pag 28 formre de1145] Section 4 Dial Pan implementation Tees Reto mue FUle 1 /A5SHN....\Y/111\1/ Structure: Matching ——_ and Replacement wasing Fenizanen Strings ny ‘omg Inthe matching string, the ‘S85 means that forthe voice translation ale to match a sting, the string must begin with 88S. Next, no- tice the Vl), which matches any four digits. Also notice that he four periods are inside ofa st of parentheses, and the parentheses are prepended with hackslashes. The backslashes ar used to negate any special meaning associated with the parentheses. The paren those themselves are used to identify a set, which can be referenced inthe replacement string. Because this is the fist and only sc in ‘the matching patter, the set number is one. In summary, fora string to match this Voice translation rule, it must begin witha SSS and ‘have at least four additional digits after the SSS. Als, those four extra digits canbe referenced by the replacement pater asset one “The replacement string begins with 11 meaning thatthe firs three digits of the translated string will begin with 111. The replace- ‘ment sting ends with a\, which refers to Sct one from the matching pater. For example, ifthe matched sting were SSSIS4S, the replacement string would be 11345, When the voice transation rule is create, it canbe referenced by a voice translation profile. Continuing wth the preceding example, ‘examine the syntax in Example 4-5 Example 4.5 Voice Translation Profile Example Router{conés)¢ voice tranelation-profile OrFIcE-coDe Router(oéq-trenslation profile)? translation called 1 In Example 45, the voice tanslation rule created in Example 4+ i eferenced by a voice translation profile named OFFICE-CODE. (sing the Voie translation rule's umber of 1) Notice the ealled keyword in the translation called 1 command. This keyword ‘means that the translation rule willbe applied toa called number (that is, a number that was dialed), “The voice translation profile then needs to be applied to an entity such as a voice por of a dal pose for it Example 4-6 shows the voice tranlation profile being applied toa voice port[148 Section 4 Dial Pan implementation Example 4.6 Applying a Volce Translation Profile Rowter(conég-voicepert)# translation-profle incoming OFFICE-CODE ‘The result ofthe command shown in Example 4 s that dialed numbers coming into the voice port that match the matching string hat is, string of numbers beginning with $55 and followed by at least four additonal numbers) willbe replaced witha seven-digit num= ber beginning with 111, with the fourth through the soventh digits copied directly from the matching string to the replacement sting, Verifying Digit Manipulation Therese a couple commands tha ar useful for verifying and testing your dil pan: 1 show diaiptan umber: This command displays the matching outgoing dal peers fora telephone number in he order they ‘would be used by the system, 1 show voice transaton-profile: This show command displays all or slected voice wasltion profiles, The formats a litle easier to red than sing the show running-config command 1 show translation ule: The ouput ofthis command displays all or elected voice anslaton rules, 1 test voice translation-re: This command is used to test the output of translation rules. The rule and number to be ested are catered afer the command. This command is very useful to et various numibr combination against auc before the rls applied to a dial perf a voice port. Influencing Path Selection Various 10S configurations can influence the route selected for placing a phone call For example, the previously discussed prefer- ‘ence dial-pecr configuration mode command can be used, In addition, tai-end hop off (TEHO) can be used to leverage the IP WAN to ‘more cost-effectively place calls tothe PSTN, Influencing Path Selection with Dial-Peer Configuration Mode Commands Earlier you read about various dial-posr configuration mode commands that you can use to match incoming and outgoing dial pers, ‘many of which enable you o influence call path selection Asa review, examine the dal-per configuration mode commands pre- sented in Table 4-5,1147] Section 4 Dial Pan implementation Table 4-5. Dial Pe ‘Regular Expression Description ‘Configuration Mode Commands for Call Routing destination-pattera Used by the ougong dial-peer matching fo match a dialed number (First choice) Matches Dialed Number Identification Service (DNIS) information for inound incoming called-number codes Ma ic Number Identification (ANI) information for inbound angwer-address ‘Third choice) Matches Automatic Number [denification (ANID information for sabowd Gal- destination-pattera peer matching Note that destination-pattern can be used for both inbound and outbound matching. Allows. a router to take the DNIS digits coming in, fom the PSTN, for example, and forward ireet-tnward-Gial ‘call based on those digits, without presenting a dial tone to the caller Breaks a ie between equally matched dial pers, where lower-preference values are more preference [0 10] pete In addition, on some ISDN links, you might be able to benefit from the no dial-peer outbound status-check pots command. This ‘command disables the checking ofthe status ofan outbound POTS dial peer during the call setup process, which might otherwise disallow a dial peer whose status was down, This might be needed when connecting to equipment that des not preset the expected signaling to the voice port Influencing Path Selection with Tail-End Hop Off Imagine a scenario where a company has two ofices, one in New York and one in Dallas. Suppose that you are the New York ‘office and want to cll one of your company’s suppliers in Dallas. You will pay toll charges if you call over the PSTN. However, be= ‘cause your office in Dallas has a PSTN gateway that can place a local call to the supplicr, you can call across the company's IP WAN, from New York to Dallas, and hop off of the Dallas PSTN gateway to make a local call 1o the supplir. This scenario uses feature called til-end hop off (TEHO)[148] Section 4 Dial Pan implementation "TEHO can be especially beneficial for corporations that have multiple geographic locations. However, consult local restrictions be- fore implementing TEHO, because TEHO isnot legal in some countries, ‘Toconfigure TEHO, follow these steps: Step 1. Create outbound VoIP dial peers to point across the IP WAN. Step 2. Perform digit manipulation (for example, to prepend the digits necessary to place a call tothe PSTN). Step’. Create a POTS dial peer at the til-end site, which points out tothe PSTN. Step 4. Perform digit manipulation at the tail-end router to modify the calling and called number to match the local PSTN, Configuring Calling Privileges ‘Torestrict calls on an 10S router, instead of using partitions and calling search spaces (as used by Cisco United Communications Manager [UCMD, you can use an approach called class ofresrction (COR), Be aware, Hower’ creation of addtional dial pers, as cmparcd to what might otherwise be required. Specifically, instead of having a single dial poer to point out othe PST, you might need to create a dal peer to point out othe PSTN for emergency calls, another dial pect to point out othe PSTN for local alls, nother one for long-distance calls, and perhaps another one for international calls. By being this granular in your configuration, you can create different rules for different categories ‘of PSTN destinations COR Theory (Class of restriction is conigued by ereating a series of COR names. These COR names are then assigned to a COR lst, These COR list ean then be assigned to incoming outgoing dal pers (ran ephone-dn in a Cisco UCME environment, of 8 Surivable Remote Site Telephony [SRST] configuration, To help understand hv COR is configured thnk of an incoming COR list (for ‘example, a COR list applied to an inbound dal pes) as containing a set of key that is, the CORS assigned tothe incoming COR Also, think of an outgoing COR lst (for example, a COR lis applied to an outbound dal pect) as containing a st of locks (that the CORs assigned to the outgoing COR lis).[149] Section 4 Dial Pan implementation oth the incoming and outgoing dial peers for a call have a COR list assigned, for the call 0 be placed, the incoming COR list must hhave a COR (hats, a key) that matches each of the COR (that is, the locks) in the outgoing COR list. cither the incoming dia ‘peer or the outgoing dal peer does not have a COR list assigned, however, thecal is permite Figure 4-7 shows an example ofa success all, The processing ofthe call goes through the fllowing steps: FIGURE 4-7, notrmg GOR List ~LOCALAN ‘neon COA List LOGAL-CUT COR Sample Topology: a = Carrernaea “one dion SmTERNAL es “9cHL do OL care cise |] @ [Gateway mucrooan |W catanay meres an | @[ ro cane tees wureerot| | reansu Potead’ | | caminapcrs sia” | [pame tSeaunreors | | ESeatour con ma sae span 4. Acallerin a VoIP network picks up a phone and dials 5851345 2. The cal is placed from an analog phone. So, when the call comes into the router, the router matches an incom- ing dial peer that points to the local FXS port of 1/00, That incoming dial peer has an incoming COR list of LOCALAIN associated with it 3. The dial string of $551345 matched an outbound POTS dial peer pointing out tothe PSTN, and that dial peer has the LOCAL-OUT outgoing COR list applied (22011 Cico Systm Io lights reserved. The publction fe protected by copyright Pisce ane page 238FIGURE 4-8 (COR Sample Topology: CallDeniea [150] Section 4 Dial Pan implementation 4. The call is permitted, Recall that for thecal to be permitted, the incoming COR list must contain all the COR names that are contained inthe outgoing ‘COR lst. In this example, the incoming COR list contains the following COR names: 911, INTERNAL, and LOCAL. Recall that you ‘can metaphorically think of COR names in an incoming COR list as keys. Also in this example, the outgoing COR list contains the following COR name: LOCAL, ‘Therefore, in this example, because the incoming COR lis contains the COR name of LOCAL (that is, the ey of LOCAL), it matches the one and only COR name contained inthe outgoing COR list, which is LOCAL (that is, the lock ‘of LOCAL). Therefore the call is permited. However, consider an example where a call would be denied (see Figure 4-8) iranag GOAL =LOSALN Yeon SOR OTF one oe “ren Leo} “aca eo © cate aise a | © [Gateway muctosan | @ Gateway matches m | @ [Tre ca Ryerss | | rewmyroread’ | | cargrersie” | [rose Puta orm tas be earths Be LD ‘se, aie ‘The process the router uses to reject the all is outlined inthe following steps 41. In this example, the caller dials a long-distance number, 2. The calls is placed from an analog phone. So, when the eall comes into the router, the router matches an incoming dial peer that points to the local FXS port of 1/0/0, That incoming dial peer has a COR list of LOCAL-IN associ- ated with it[51] Section 4 Dial Pan implementation 3. The dial string of 8595551345 matched an outbound POTS dial peer pointing out to the PSTN, and that ial peer hhas the LD-OUT COR list applied 4. The call is denied is example, the LD-OUT COR list contain the COR of LD. However, because the incoming COR list of LOCAL-IN does not contain the COR of LD, th calli rejected, Special Conditions [no COR list statements are applied to some dial pects, the following properties apply © When no incoming cortist command is configured ona dial pec, the default incoming COR lis is used. The default ihest possible priory, and it therefore allows this dial per to acess al ther dial pee 9 their outgoing COR list When no outgoing corlist command is configured ona dial pos, the default outgoing COR list is used. The default outgoing ‘COR list has the lowest possible priority, and it therefore allows al other dal peers to access this dal peer, regardless oftheir COR ist COR Configuration ‘The frst step when configuring COR is to configure COR lists, which is performed in COR custom configuration mode, as demon- strated in Example 4-7. Ifyou ae familar with creating partitions and calling search spaces in a UCM environment, this stp is some- ‘what analogous tothe creation of partitions. In this example, four CORS are created: 911, INTERNAL, LOCAL, and LD. Example 4.7 Creating CORS Rogter{conls)¢ dial-peer cor eu Roater(conss-d Router conga Roster oons3-a Rover (oonss-d Next, COR lists are exeated. ACOR list isa collection of CORS, and these COR lists can be applied to incomingloutgoing dial peers. Example 4 illustrates the configuration ofthe incoming COR lists uscd in this example. Four COR tobe used as incoming[152] Section 4 Dial Pan implementation ‘COR lists, ar create inthis example: 911-IN, INTERNAL-IN, LOCAL-IN, and LD-IN. Notice that the member command is used toadd CORS to COR list. These COR were the keys used inthe prvious metaphorical explanation Example 4.8 Creating Incoming COR Lists Roxter{conig)¢ diai-peer cor eu Rogter(consg-dp-corlist)# member 912 Rogeer{oonsg-dp-corlist)# exit Rogeer{conss)¢ dial-poer cor ou Router{coniq-dp-corliet)¢ menber 912 Router{conég—dp-corlist}# member THEERKRE Router conss-dp-corlist)# exit Rogeer{conss)¢ dial-poer cor ou Router oonsg-dp-corliet}# member Rogterconss-dp-corlist)# member Rogeer{conss)¢ dial-poer cor ou Rogter(consg-dp-corlist)# member 912 Router{conig—dp-corlist}# member THEERKRE Rovter{conig—dp-corlict}# member LOCAE Rogeer{consg-dp-corlist)# member 2D Example 49 shows the eeation ofthe outgoing COR lists used inthis example, Members of these COR lists are the locks that need tobe unlocked bythe keys inthe incoming COR lis. Example 4.9 Creating Outgoing COR Lists Roter(conss)¢ Roster {consg-a Rovter(conig-dp-corlist}# exit Rogter{conSs)¢ dial-peer cor cu Router{conig-dp-sorlist)¢ member LOCAL[153] Section 4 Dial Pan implementation Roster {conég-dp-corlist)# exit Rogeer{conss)¢ dial-peer cor ou Rogter(conSg-dp-corlist)# member 2D ‘After the creation ofthe incoming and outgoing COR incoming COR listo an existing POTS dial peer 5 need to be applied, Example 4-10 shows the application of an Example 4-10 Applying an Incoming COR List Router{conSa)¢ dial-peer voice 100 pots Rogeer{consg-dial-pesr)# corlist incoming LOCA Finally, outgoing COR lists are applied to outgoing dial poe. To illustrate, Example 4-11 shows the application of an outgoing COR list o.an outgoing dial pect. Example 4-11 Applying an Outgoing COR List Roster{conSs)¢ diai-peer voice 200 pote Rogeer{oonsg-dial-pesr)# corlist oxtgoiag LD-OUr Using COR can benefit not only POTS and VolP dial poets but also UCM Express (UCME) and SRST configurations. Both UCME and SRST configurations allow Cisco IP Phones to register wih a router running the UCME or SRST feature. COR can therefore be used to enforce call restctions on those Cisco IP Phones. Specifically the eo fincoming| outgoing} ist-name command can be used in ephone-dn configuration mode on a UCME router, andthe cor fincoming | outgoing} lest-name tag starting umber — _ending-rumber command can be use incall manager fallback configuration mode 1 apply a COR listo arange of SRST directory ‘numbers Verifying COR You can verify COR settings withthe following four commands show dial-peer cor endpoints. {= show dial-peer voice: This command displays the parameters ofa voice dial peer, including the incoming and outgoing COR list, ‘This command displays the COR names (labels) and lists applicable to alltypes of dial peers and[154] Section 4 Dial Pan implementation show voice register pool: This command displays the parameters ofa SIP endpoint, including the incoming and outgoing COR ist. show running-config | begin ephone-dn: This command displays the configuration of SCCP ephone ns, including the incoming and outgoing COR list. Because there is no SCCP-speciic command for display COR parameters, you must view a part ofthe gateway configuration (22011 Cico Systm Ie lights reserved. The publlction fe protected by copyright Pisce ane page 28 formare detail:1155] CNP Voice QVoice 642-437 Quick Reference Section 5 Gatekeeper and Cisco Unified Border Element Implementation ‘An H.323 gatekeeper offers multiple benefits for larger enterprise networks. For example, a gatckeeper can make sur thatthe band wii inthe IP WAN does not become saturated with too many simultancous voice calls. This section investigates this and many her features ofthe gatckeeper, in addition to exploring the theory and syntax ofa basic gatekecper configuration. You will also see how to scale your network with multiple gatekeepers, Understanding Gatekeepers ‘Among a gatckesper’s many feature, some are mandatory and some are optional. This section distinguishes between the two, In ad- dition, gatekeepers use a series of registration, admission, and status (RAS) messages, and some ofthe most commonly used RAS messages willbe described, You will alo see how directory gatkeeprs can help scale a network along with the use of technol- ogy prefixes Finally, you wl lear that you can ofload some of gatckssper's processing burden to an external sever using the Gaickscper Transaction Message Protocol (GKTMP). Gatekeeper Platforms Cisco 10S Gatekceper isa software festre that rns om various Cisco outers, A special IOS image is equted in Integrated Series Routers, Generation | and license keys in Integrated Services Rowers, Generation 2. The following is list ofthe curently supported router platforms that support Gatekeeper © Cisco ASS3S0XM and Cisco ASS400XM Universal Gateways[156] Section & Gatekeeper and Gisco Unified Border Element implementation Cisco 2800 Setes Integrated Services Routers, Gl Cisco 2900 Series Incgrated Services Routers, G2 © Cisco 3800 Series Integrated Services Routers, Gi Cisco 3900 Series Integrated Services Routers, G2 Cisco ASR 1000 Series Aggregation Services Routes “The routers will also need a special version ofthe IOS to support Gatckeeper. IOS version 124 and carler require one of the follow ing feature sos: = INTVOICE/VIDEO, PIP Gw,TDMIP GW = INTVOICE/VIDEO, IPI>GW, TDMIP GW AES = INT VOICE/VIDEO GK, IPIPGW, TDMIP GW AES, LI 10S version 15 uses a universal IOS image. Gatekeeper is licensed process within the router. That licensing i enforced withthe use of license files stating in IOS Version 124(20)T Gatekeeper Functions Even though a gatekeeper isnot a required component in an H.323 network, a gatckeeper could certainly help an H323 network to scale, in addition o performing a collection of other features. Some ofthe gaickeeper’s features are mandalory, whereas others are ‘optional, as shown in Table 5-1 (©2011 Cico Systm Ie lights recered. The publloton ie protected by copyright Piso sn page 228 for mare dtlTable 5-1 Gatekeeper Features 1157] Section & Gatekeeper and Gisco Unified Border Element implementation Feature | Mandatory or Optional _| Description ‘Address tana tea dialed phone number into a comre- Address translation Mandatory sponding IP address to which a setup packet should be sent ro ton ‘Admission control uses a series of RAS messages to pemmit or deny Admission control Mandi saree oa Bandwidth contol Mandatory ‘Bandwidth control supports midcall codec negotiation, ‘Zone management can provide features (for example, address transla- Zone management Mandatory ion, admission contro, and bandwidth control) 10 a device based on the logical zone with which a terminal or gateway is associated authorization tion Call authorization can permit or deny access o H 323 endpoints based Call athorizati Optional on policies (for example, time-of-day policies). Call management Opsonal Call management keeps status information on current calls and infia- ences routing of additional calls based on that information. ‘Bandwidth management | Optional ‘Bandwidth management is subset ofthe admission control feature, and can permit or reject acall based on available bandwidth. Call contol signaling | Optional ‘Even though 235 and H1235 cll control signaling messages are ex changed directly between the H.323 endpoints involved ina call, the ‘gatekeeper supports a configuration where these protocols can low through the gatekeeper “Tovndestand the thor and operation of gatsheeper, you ned to understand fe basic trms. These emt he ollowing: one, sone pfs, and chology pai: Definition of a Zone As mentioned in Table 5 to which they belong. tckeeper (GW) can treat H323 endpoints (that is, terminals or gateways) differently based onthe zane igure 5-1 shows a simple topology with one gatekeeper,[158] Section & Gatekeeper and Gisco Unified Border Element implementation FIGURE 5-4 Gatekeeper Locat ones ox Zarek en [Notice that each H.323 GW belongs toa zone, but the gatceeper tse doesnot belong toa zone. When cach of these gateways resis: ters withthe gatekeeper, it registers as member ofits zone Because members of both ZoncA and Zone® regiscr withthe gatekecpe, ‘both zones ar considered t be fcal zones, fom the perspective of the gtekexper. Ifthe topology contained an addtional gatekeeper, as illustrated in Figure 5-2, and Zone had registered with gachecper GK2, “ZoneB would be considered tobe a remot zone from the perspective of gatekeper GK, However, ZoneB would be considered o be local 20a fom the perspective of gatkeeper GK2. FIGURE 5.2 Gatekeeper Remote Zones ‘ot one Zonk Zone (22011 Cico Systm Ie lights reserved. The publlction fe protected by copyright Pisce ane page 28 formare detail:[159] Section & Gatekeeper and Gisco Unified Border Element implementation Definition of a Zone Prefix ‘When a gaeeper leans about the H.323 terminals and gateway that have registered as members of various zones, the gatekeeper ‘an direct calls destined for particular phone numbers toa terminal or gateway ina zone tht can appropriately route a call destined for that phone number (see Figure 53). FIGURE 5-3 Gatekeeper Zones, Prefixes ane Prekc508 Tone Pre 359 In Figure 5.3, GW registered with GK1 as a member of ZoneA. GW2 registered with GKI. asa member of ZoneB. The gatekeeper (GK might be configured with zone prefixes that say’calls destined for rca code 606 shouldbe forwarded to ZoneA, and calls des- tined for area code 889 should be forwarded to ZoneB. Asa result, if 606-585-1111 calls 859-588-2272, gateway GWI will sk gate keeper GK. how to route the call. Gatekeeper GK1 will examine the dial string and notice the destination area code of 859, The gackeeper's zone prefix configuration states that if a cll is destined for area code 859, it should be roued to ZoncB, Because gateway GW2 has rogistered with gatekeeper GK1 as a member of ZoneB, gatekeeper GK will tll GWI that t should send call setup message directly to GW2. Definition of a Technology Prefix ily physician wh handles mos of yourmedcal neds and regular cheskups. On aeasion, however, you might 1 For camp, you might go toa dermatologist for your skin a podiatrist fr your fet, or an ophthalmologist for your yes. Similar, when a gateway registers wih patskeper, the gtenay can reser as a specialist fora parculartype of call,[160] Section & Gatekeeper and Gisco Unified Border Element implementation for video calls for modem calls, for fax calls, or for another specific call ype. a default technology gateway. A default technology gateway can For example, a gnteway might registra pes Altematively, much lke a family physician a gateway can regis handle calls that id not request a pscai. ‘The way a call requests a specialist is by specifying a technology prefix as part of the dil string. When the gatckeeper receives a dial string that is prefixed witha technology pr‘, the gatekeeper can forward th call to @ gateway that registered with that specific tech- nology prefix. Registration, Admission, and Status Gatekeepers communict with H.323 endpoints and gateways, in aditonootbergalckeeprs, using RAS messages. Alboughsev- eral RAS messages exit, these reference sets describe some ofthe more commonly sed RAS messages, Discovery RAS Messages Before an H_323 terminal or gateway can register with a gatekeeper, it nceds to discover a gackeeper. Pethaps a gateway is preconfig- ured with the IP address ofa gatcheeper. In that case, the gateway sends a gatckecper request (GRQ) RAS message to make sure that the gatckecperis up and responsive. This GRO is sent using a unicast message directed foward the preconfigured IP address, However, i the gateway docs not know the IP address of gatckeeper, ican attempt to dynamically discover a gackeeper. This dy- ‘namic gatekecper discovery involves sending a GRQ message toa multicast addres of 224.0.141 using UDP port 1718, 1 gatckseper confirm (GCF) RAS message to say that itis up and available. If “however, it can respond with After a gatcheoper receives a GRQ,itcan respond wi the gatckeeper isnot availabe (that is, it does not want the endpoint to attempt to register 8 gatekeeper reject (GRJ) RAS message. Registration RAS Messages ‘When an endpoint has discovered a gatekeeper, the next step is for that endpoint to register with that gatekeeper. Specifically, the ‘endpoint sends a repistration request RRQ} RAS message tothe gatckecper requesting permission to register using UDP port 1719. As part of this registration message, the endpoint tells the gatekeeper about itsel. For example, an RRQ RAS message might include information such as the name ofthe endpoint, the IP address of the endpoint, and the phons numbers reachable by that ‘endpoint