This action might not be possible to undo. Are you sure you want to continue?
AT&T Drive Test Results and Report Preview
MCS SNR MIMO RI BLER 64QAM 16QAM QPSK CQI L1 DL RSSI RSRQ RSCP CINR ACK NACK PDSCH HS-PDSCH HSSCCH MAC-HS E-DPCH FTP UDP HTTP RTT TCP dBm dB Redefining Research Mbps ms MOS LTE HSPA+ DC-HSDPA TD-LTE EV-DO Rev A/B Mobile WiMAX MCS MCS SNR MIMO RI BLER 64QAM 16QAM QPSK CQI L1 DL RSSI RSRQ RSCP CINR ACK NACK PDSCH HS-PDSCH HS-SCCH MAC-HS E-DPCH FTP UDP HTTP RTT TCP dBm dB Mbps ms MOS LTE HSPA+ DCHSDPA TD-LTE EV-DO Rev A/B Mobile WiMAX MCS MCS SNR MIMO RI BLER 64QAM 16QAM QPSK CQI L1 DL RSSI RSRQ RSCP CINR ACK NACK PDSCH HS-PDSCH HS-SCCH MAC-HS E-DPCH FTP UDP HTTP RTT TCP dBm dB Mbps ms -50 -40 -30 -20 MOS LTE HSPA+ DC-HSDPA TD-LTE EV-DO Rev A/B Mobile WiMAX MCS MCS SNR MIMO RI BLER 64QAM 16QAM QPSK CQI L1 DL RSSI RSRQ RSCP CINR ACK NACK PDSCH HS-PDSCH HS-SCCH MAC-HS E-DPCH FTP UDP HTTP RTT TCP dBm dB Mbps ms MOS LTE HSPA+ DC-HSDPA TD-LTE EV-DO Rev A/B Mobile WiMAX MCS MCS SNR MIMO RI BLER 64QAM 16QAM QPSK CQI L1 DL RSSI RSRQ RSCP CINR ACK NACK PDSCH HS-PDSCH HS-SCCH MAC-HS E-DPCH FTP UDP HTTP RTT TCP dBm dB Mbps ms MOS LTE HSPA+ DC-HSDPA TD-LTE EV-DO Rev A/B Mobile WiMAX MCS MCS SNR MIMO RI BLER 64QAM 16QAM QPSK CQI L1 DL RSSI RSRQ RSCP CINR ACK NACK PDSCH HS-PDSCH HS-SCCH MAC-HS E-DPCH FTP UDP HTTP RTT TCP dBm dB Mbps ms MOS LTE HSPA+ DC-HSDPA TD-LTE EV-DO Rev A/B Mobile WiMAX MCS MCS SNR MIMO RI BLER 64QAM been there, done that… 16QAM QPSK CQI L1 DL RSSI RSRQ RSCP CINR ACK NACK Includes additional information about our forthcoming PDSCH HS-PDSCH HS-SCCH MAC-HS E-DPCH FTP UDP series of network technology benchmark reports HTTP RTT TCP dBm dB Mbps ms MOS LTE HSPA+ DCHSDPA TD-LTE EV-DO Rev A/B Mobile WiMAX MCS MCS
8,000 kbps 7,000
Other Other Nokia Kyocera Palm LG
100% 90% 80% 70%
Septemer 18, 2011, Vol. 7 No. 10
HSPA Demo Phone - Cat 8 (1Rx)
Total DC-HSDPA PHY Layer Throughput (Kbps)
2,000 1,000 0
PHY Layer Primary Throughput (Kbps) PHY Layer Secondary Throughput (Kbps)
36:45.0 36:57.0 37:21.0 37:45.0 37:57.0 37:09.0 37:33.0 38:09.0
HSPA Demo Phone - Cat 8 (1Rx)
20% 10% 0%
17:29.0 17:37.0 17:45.0 17:53.0 18:01.0 18:09.0 18:17.0 18:25.0 18:33.0 18:41.0 18:49.0 18:57.0 19:05.0 19:13.0 19:21.0 19:29.0 19:37.0 19:45.0 19:53.0 20:01.0 20:09.0 20:17.0 20:25.0 20:33.0 20:41.0 20:49.0 20:57.0 21:05.0 21:13.0 21:20.0 21:27.0
based on some of the more frequently cited key performance indicators (KPIs). including the relative differences in cell site density.2% of the time. thus providing invaluable competitive intelligence. In fact. In collaboration with Accuver. by testing in the Houston market on two different LTE networks we were able to compare and contrast the performance of the two infrastructure suppliers who have the preponderance of the LTE market share in North America.0 Executive Summary AT&T just launched commercial LTE services in Dallas.6Mbps with a peak data rate of 61. It goes without saying that we tested all of the networks in the Houston area and that we contained great insight on how they perform. Number 10 . we happened to find ourselves in Houston prior to the This special-edition summary report is provided on a complimentary made available to other interested parties. Coincidentally. Chicago.6 miles of the greater Houston area. and San Antonio using its 700MHz spectrum assets. we are using this summary report to promote a series of forthcoming. we encourage you to do so. a leading supplier of drive test equipment. we offer the following observations about the performance of AT&T’s LTE network. Unlike our regular Signals basis to our existing Signals Ahead corporate clientele and it is being launch where we had the chance to kick the tires of the operator’s network. reports in which we will take our network benchmarking capabilities to a whole new level. In addition to quantifying the performance of the network. Houston. recipients of this report may feel free to forward this report to friends. far more expansive. Atlanta. Both results meaningfully exceeded our expectations. Arguably.6% of the time and 21Mbps – the theoretical peak data rate of the operator’s HSPA+ network – for 38. is contained later in this report. by carefully selecting the markets in which we conducted our tests. ■ The average downlink Physical Layer throughput was 23. Vol. air interfaces. The data rate also exceeded 40Mbp for 8.1Mbps. we were able to include virtually every single combination of infrastructure vendor and technology. and after covering 205. and/or devices in a seamless fashion. which can impact performance in the near term and inherent capacity in the long term. the average data rate would have been even higher if we included a mix of stationary testing from optimal locations and/or included results from pedestrian testing. we have been using its XCAL-MO network benchmarking tool.8GB of data over a three-day period with the Sierra Wireless “USBConnect Momentum 4G” dongle. Most importantly. colleagues and business associates – in fact. Our approach. By testing in the Houston market on two different LTE networks we were able to compare and contrast the performance of the two infrastructure suppliers who have the preponderance of the LTE market share in North America. the data rate was greater than 5Mbps for 95% of the time. 2 September 18. 7. Based on transferring 88. Ahead research. 2011 | Signals Ahead. which is strictly limited to paid subscribers. which was specifically designed to facilitate the simultaneous testing of multiple networks. which is included as part of a normal Signals Ahead subscription.1. Additional information about this series of reports. Further.
the HSPA+ throughput following the handover was quite good (median MAC Layer throughput = 5. on at least one occasion we hit a data rate in excess of 60Mbps while traveling at 54mph (reference Figure 8). Over time.1Mbps) was achieved while traveling at 64mph. Transitioning to wider channel bandwidths (e. 3 September 18. That being said. the latency was on par with other LTE networks in the North American market. This result was higher [worse] than we were expecting based on the capabilities of LTE and our experiences with testing networks in Europe. was 49ms with a minimum value of 40ms. impact the perceived network performance. and for most applications it would have more than sufficed. but are not limited to the following: ➤➤ The use of a sophisticated drive test tool which can capture accurate. Vol. including the effects of the network and any transport and IP routing delays which may have existed. That being said.5 seconds. ■ Intra-RAT handovers from LTE to HSPA+ were relatively seamless for a typical data application. The handover time is based on the device dropping the LTE connection and establishing a connection on the HSPA+ network using the CELL_DCH state. we surmise that most users would prefer to be on the higher throughput/lower latency LTE network. Sixty percent (60%) of the time the uplink Physical Layer data rate exceeded 15Mbps and 98. Number 10 . in particular in the infancy of LTE. and a user will probably never notice the interruption. applications. That being said. ■ These results are specific to AT&T’s LTE network in Houston and our test methodology. but since LTE is not currently being used to support voice services there is still ample time to lower the handover latency. will require a much shorter handover time. Worth noting. ■ The average latency. we used the File Transfer Protocol (FTP) to transfer multiple files simultaneously to a server located in a third-party Cloud-based server that we leased in the Chicago area as well as to FTP servers in Dallas and Phoenix. Obviously. such as voice. We had no expectation on what these values should be but from our perspective these values are quite low. the average latency to this server using the Verizon FiOS service in the Dallas area was 14ms and we observed latency as low as 27ms in the Houston area to this server while using a commercial 3G network during the morning rush hour. is far more statistically significant than stationary testing from a couple of random locations. optimized routing and tighter integration with the operator’s wireline assets and IP backbone can help reduce the latency outside of the network. we also know that several factors can. Handovers from HSPA+ to LTE currently require the device to return to the IDLE state. These results are based on two instances where our device handed off to the 3G network. ■ The average uplink Physical Layer throughput was 15. requiring 2. The uplink results were also much higher than we anticipated. 2011 | Signals Ahead. 20MHz channels) can also reduce the latency in the radio access network. An HSPA+ device uses the dedicated channel (DCH) when data is being transferred on an HSPA+ network.6Mbps. Similar to the downlink throughput testing. The uplink throughput shown in Figure 9 (23. real-time information.. These factors include.2Mbps with a peak data rate of 23. and do.4 seconds to exit the LTE network and establish a dedicated connection on the HSPA+ network.however. In our testing. 7. Although we have no reason to expect materially different results in other AT&T markets.9Mbps). CELL_DCH is one of the RRC (Radio Resource Control) states used by an HSPA+ device. which could result in highly erroneous results – either positive or negative. The total time during which the data transfer was interrupted was 6.2% of the time it exceeded 5Mbps.g.
From what we have seen and documented. and chipset supplier. Number 10 4 . even with the use of sophisticated drive test tools. pedestrian and stationary testing. We also include a Test Methodology section. which focuses almost exclusively on vehicular testing. to the extent we did detect other people accessing the network. September 18.. Over time the performance capabilities and features of each important element should converge. device manufacturer. but that there are relatively so few of them today and the combination of their distribution across the network and their usage patterns tends to minimize or completely eliminate the impact of their presence. and that [at least some] operators are establishing tariffs that curtail bandwidth hogs.➤➤ The combination of infrastructure supplier. Without question. they will never be achieved. Further. 7. in the early days of LTE there are measurable performance differences that exist. topography of the market being tested. ➤➤ Access ➤➤ The topology of the radio access network and the use of a distributed network architecture versus a traditional macro cellular network. ➤➤ The ➤➤ The densification of the network cell sites. Vol. as well as the user experience metrics. We will be doing this research in collaboration with Spirent Communications. In a pending Signals Ahead report we will be examining the performance differences that exist with today’s LTE chipsets. in particular when testing very high bandwidth networks. as well as how the operator’s cellular network interfaces with its IP backbone and the location of its peering points with the Internet. Following the Introduction. ➤➤ The ➤➤ The ➤➤ The backhaul architecture. tends to understand the true capabilities of the network. popular web testing portals may understate the throughput. Further. 2011 | Signals Ahead. but this convergence does not exist today. we welcome the chance to speak with any organization that is interested in using our services and the Accuver suite of benchmarking tools to conduct benchmark studies that are specifically tailored to a particular set of customer requirements. and information about how you can pre-order this series of reports. A denser deployment of cell sites should deliver higher ■ It will be quite some time before network loading has a meaningful impact on the throughput for a typical user. Our methodology. the results that these services provide represent an average as measured over a very short period of time during which only modest amounts of data are actually transferred. This observation does not suggest that there are not any LTE subscribers and that they are not using the network. there wasn’t a material/long-term impact on the results we were obtaining. the Detailed Results section provides the throughput and latency results for all of the AT&T LTE test scenarios from the greater Houston area – exclusive of the results which involve other technologies and networks. All things being equal. 10-20MHz channels) to LTE. We have done more than our fair share of drive testing in various LTE networks and we have seldom detected the presence of other users. It also doesn’t hurt that LTE is a very efficient technology when it comes to handling data and signaling traffic. placing the radio electronics on top of the tower and closer to the antenna mast will result in a higher throughput and better coverage. Unless the end-to-end network is capable of supporting the peak data rates of LTE. Bespoke Services Additionally. throughput if it has been properly tuned and optimized. to a high-bandwidth server and/or servers with sufficiently low latency in order to fully load the pipe. its throughput capabilities. some closing remarks. mix of vehicular. that operators are committing a fair amount of spectrum (e. processing capabilities of the host device and its associated operating system.g.
Thelander (510) 338 1284 mike@signalsresearchcom www. either in whole or in part. “Oh the places we did go!” (Downtown Houston) – Geo-plot of test routes with speed (mph) ………………… 12 Figure 2.0 Executive Summary………………………………………………………………………………………………………………………………………… 2 Bespoke Services…………………………………………………………………………………………………………………………………………………4 2. If you value the information and insight that we provide then I strongly urge you to respect our hard work and livelihood and subscribe to our research.4 Network Latency Results…………………………………………………………………………………………………………………………… 18 3.2 Uplink Results……………………………………………………………………………………………………………………………………………… 15 3. Vol. 7. 2011 Michael W. to the “SRG Wall of Shame.0 Test Methodology……………………………………………………………………………………………………………………………………… 20 5. If you or your organization is interested in distributing this report to outside organizations. with logo.signalsresearch. Number 10 . No. AT&T Houston LTE Network Vehicular Mode – Geo-plot of Physical Layer Downlink Data Rates…………………… 14 Figure 4.0 Detailed Results…………………………………………………………………………………………………………………………………………… 11 3. AT&T Houston LTE Network Latency Results – Pie Chart Distribution…………………………………………………………… 18 Volume 7. “Oh the places we did go!” – Geo-plot of test routes with speed (mph)……………………………………………………… 11 Figure 1A. 10 September 18. please contact us for permission and we will be happy to provide it.Contents 1. 2011 | Signals Ahead. If you forward this research to external organizations. you may want to upgrade your license so that you can share this issue across your entire organization with our blessing. We also reserve the right to post your company’s name.0 Conclusions………………………………………………………………………………………………………………………………………………… 24 Index of Figures Figure 1. 5 September 18.com WITH THE EXCEPTION OF THIS ISSUE. or if you share the contents of the report beyond the authorized allocation within your organization then the licensing terms for our research are being violated.5 The Statistical Significance of 90GB…………………………………………………………………………………………………………… 19 4. AT&T Houston LTE Network Uplink PHY Layer Throughput Results – CDF and Pie Chart Distribution…………… 15 Figure 5. AT&T Houston LTE Network Downlink PHY Layer Throughput Results – CDF and Pie Chart Distribution………… 13 Figure 3. If you do not have a platinum license or a global license.0 Introduction……………………………………………………………………………………………………………………………………………………6 Volume 1 (Network and Technology Performance)……………………………………………………………………………………………… 7 Volume 2 (Quantifying the User Experience)………………………………………………………………………………………………………8 Volume 3 (Detailed Performance Analysis)……………………………………………………………………………………………………………9 3.” If you received this issue from someone outside of your organization and it did not come directly from SRG then the licensing terms for our research are being violated. If you would like to leverage a figure or quote from this report and you have at least a global license. please feel free to contact us to discuss licensing terms and fees. AT&T Houston LTE Network Vehicular Mode – Geo-plot of Physical Layer Uplink Data Rates……………………… 16 Figure 6. any unauthorized use of our research material will result in the non-refundable cancellation of your subscription. Inter-RAT Handover………………………………………………………………………………………………………………………………… 17 Figure 7.3 Inter-RAT Handover Results………………………………………………………………………………………………………………………… 17 3.1 Downlink Results………………………………………………………………………………………………………………………………………… 13 3.
even this level of information far surpasses anything that we have come across in the public domain. Signals we have grown quite fond of their capabilities and their ease of use. offering its performance government regulators. thus turning a multi-person task into a relatively simple chore that a single individual could perform. For the first time. The first volume of the report. count is a poor proxy for value. and Samsung infrastructure suppliers who (Mobile WiMAX). Vol. which counts as two Signals can be purchased on Ahead issues. yet from our perspective it barely scratches the surface of what we have in store in the yet-to-be published reports. As a preview to this forthcoming series of reports we are providing a sneak peek into the performance of AT&T’s LTE network in Houston. Although page an individual basis. Operators in this study include AT&T (HSPA+ and LTE). Charleston. San Francisco. We have used the Accuver tools several times in the past for various Signals Ahead reports and In collaboration with Accuver. We benchmarking services to interested organizations believe that the powerful combination of their tools along with our unique perspective and ability on a global basis.0 Introduction For the last few months we have been collecting network performance data in a number of US markets. HSPA+ and Mobile WiMAX). and [most recently] Houston. trade associations or other interested parties on a global basis. Ericsson (DC-HSDPA.2. we will have included virtually every commercially-available next-generation broadband wireless technology known to mankind. When all is said and done. Myrtle Beach. The culmination of this independent study is a series of three reports. dubbed “The Mother of all Network Benchmark Tests – How Not to Spend Your Summer Vacation. Oakland. mobile operator network benchmarking. is largely finished and we expect to be releasing it in late September. A summary of each report follows in the subsequent paragraphs. reports is included Our three-part series of reports is included with a subscription to Signals Ahead or it can be with a subscription to purchased on an individual basis – the former option is far more economical since it includes at Signals Ahead or it least 14 additional Signals Ahead reports. Huawei (LTE and Mobile well as a fairly thorough representation of the major WiMAX). and Verizon Wireless (EV-DO generation broadband Rev A and LTE). 6 September 18. Clearwire (LTE 2x20MHz and Our benchmark study includes virtually all next. in its present format we estimate that the final product will consist of 125-150 pages with more than 130 figures and/or tables. Infrastructure suppliers in this study include Alcatel Lucent (EV-DO Rev A.” which we will be publishing starting later this month and continuing throughout the month of October and into November. With the exception of mobile operators who provided us with free and unlimprovide these solutions. who allowed us to use its suite of network drive test tools and the XCAP post-processing software. T-Mobile USA (DC-HSDPA and HSPA+). 7. vendors. HSPA+ and LTE). Number 10 . San Jose. be they studies involving technology/vendor comparisons. we are now Research Group is now able to offer their tools with our services for commissioned-based projects on behalf of operators. as well as a fairly thorough representation of the major infrastructure suppliers who provide these solutions. including Dallas. EV-DO. With its recently announced XCAL-MO network benchmark tool we were able to collect network performance data simultaneously on all of the networks that we tested. We look forward to discussing such opporOur three-part series of tunities with anyone that is interested. while also navigating the city streets and drinking Peet’s coffee along with munching on a bran muffin. Our ability to collect and analyze the network performance data would not have been possible without the support of Accuver.Mobile WiMAX). this study was entirely funded by Signals Research Group without any involvement from the aforementioned companies. 2011 | Signals Ahead. Nokia Siemens Networks (DC-HSDPA. as an independent research consultancy and industry analyst organization to independently validate the performance characteristics and to articulate their implications. From what we have observed in the past. or mobile device/chipset performance offers a powerful and differentiated service. Phoenix. wireless technologies as HSPA+ and LTE). Kansas City. ited access to their networks.
2011 | Signals Ahead. Further.Volume 1 (Network and Technology Performance) Next-generation network technologies are not created equal. but only one operator can be right. the following: ➤➤ Application ♦♦ ♦♦ Volume 1 is critical for operators around the globe who are currently making strategic decisions regarding their network technology evolution. Everyone claims that they have the best network. median and CDF plots Geo-plots of throughput for all test scenarios using Google Earth comparisons. and/or Physical Layer Throughput Mean. As operators around the globe struggle to make crucial strategic decisions regarding their network technology evolution. but also to the underlying technology itself. including DC-HSDPA versus LTE (2x10MHz) with 2x2 MIMO LTE (2x20MHz) versus LTE (2x10MHz) Mobile WiMAX versus HSPA+. LTE and DC-HSDPA DC-HSDPA versus HSPA+ EV-DO versus LTE. Vol. but are not limited to. 7. While this report is intended to address the needs of operators worldwide by focusing on the performance of the technologies. as a secondary feature it also provides valuable insight into the performance of each major network deployment in the United States. Number 10 . some of which are due to channel bandwidth considerations. there exist meaningful performance differences. etc Throughput normalized for channel bandwidth and duplex scheme Does LTE with MIMO really outperform narrow bandwidth solutions LTE network performance with multiple devices DC-HSDPA and HSPA+ devices in the same 10MHz channel allocations Downlink throughput Uplink throughput Latency Variance based on time of day and network loading ➤➤ Technology ♦♦ ♦♦ ♦♦ ♦♦ ♦♦ ➤➤ Single User Spectral Efficiency Results ♦♦ ♦♦ ♦♦ ♦♦ ➤➤ Side-by-Side operator network coverage maps for drive routes used in each market ♦♦ ♦♦ ➤➤ Network ♦♦ ➤➤ LTE network deployment philosophies (LTE cell site density relative to the legacy network) and their implications for coverage and capacity AT&T HSPA+ versus AT&T LTE ♦♦ ♦♦ ♦♦ Verizon Wireless EV-DO versus Verizon Wireless LTE AT&T LTE versus Verizon Wireless LTE WiMAX (2500MHz) versus LTE (700MHz) ➤➤ Mobile 7 September 18. Specific topics addressed in Volume 1 include. it is imperative that they fully understand and appreciate the potential of these technologies as well as their limitations. operator deployment philosophies and the maturity of the solutions can have an over-arching impact on the results. On paper. HSPA+.
Instead. but also for mobile operators. but are not limited to. the user experience based on HTTP web page download times Popular websites. In addition to mobile operators. based on device/chipset signaling messages Network/technology comparisons Comparisons based on network loading – same location over a 12-15 hour period of time ➤➤ Determining ➤➤ Determining ♦♦ if perceived differences in network/technology performance have more to do with network loading than the actual capabilities of the network/technology itself how the combination of throughput and latency impact the HTTP web page download time results 3 axis plot. Number 10 . But to what degree do these relationships provide the most benefit to the user? Mobile video. not only for consumers.Volume 2 (Quantifying the User Experience) Although mobile operators. and their impact on the user experience 8 September 18. it is readily apparent that the capabilities of these next-generation networks frequently exceed the requirements of the application and/or the capabilities of the Internet itself. especially if the performance of next-generation networks fails to keep up with the requirements of the video content. CNN. industry pundits and most well-informed consumers understand the notion that a higher megabit-per-second throughput is preferable. the typical consumer is generally clueless when it comes to understanding what these obscure marketing messages really mean for the mobile Internet experience. how they vary as a function of loading. including Yahoo. but it is easier said than done. 2011 | Signals Ahead. and video hosting services. Vol. it may actually be the combination of relatively high throughput and low network latency – offset by transport latency – that really defines the user experience. this report provides invaluable insight to application developers and content providers who require a greater appreciation for how network performance characteristics impact the user experience. be it YouTube or Netflix. This report is critical for operators trying to understand how to market their broadband wireless service offering as well as how they should prioritize their network optimization activities in order to achieve the best possible user experience for their subscribers. The perceived quality of the video playback also matters. content owners. iTunes. etc. this issue becomes even more important. is driving mobile data growth and the capabilities of nextgeneration networks will only serve as an impetus to even higher data usage. As higher resolution video formats with higher encoding rates become more mainstream. 7. Results down to the millisecond. Amazon. showing maximum achievable throughput and network latency versus webpage load time and required throughput Which matters most – latency or throughput Does DC-HSDPA really offer a quantifiable benefit over HSPA+ Determining the crossover point when higher throughput become irrelevant the user experience based on downloading Google email attachments the user experience based on downloading video and audio content from iTunes ♦♦ ♦♦ ♦♦ ➤➤ Quantifying ➤➤ Quantifying ♦♦ Determining the crossover point when higher throughput become irrelevant video streaming requirements ➤➤ Netflix ➤➤ Determining the chokepoints in the network (from end user to the original source of the content). YouTube. Very few applications and/or web site servers support high double-digit megabit-per-second throughput. Most operators recognize that they need to move away from the “my pipe is bigger than your pipe” marketing mentality. Further. Specific topics addressed in Volume 2 may include. the following: ➤➤ Quantifying ♦♦ ♦♦ ♦♦ ♦♦ Volume 2 is critical for operators trying to understand how to market their broadband wireless service offering as well as how they should prioritize their network optimization activities.
Number 10 . by peeling back the layers of the proverbial technology onion it becomes possible to gain a greater appreciation for how each technology delivers its results. but not limited to the following: ➤➤ MAC-HS ➤➤ MAC-HS ➤➤ Reported Throughput versus RSCP scatter plot Throughput versus Reported CQI Values scatter plot Throughput versus Cell ID (real time) Throughput versus # of Assigned HS-PDSCH Codes (real time) Throughput (primary. 2011 | Signals Ahead.. handover performance) CINR versus RSSI (scatter plot and real-time) ➤➤ Throughput ➤➤ Throughput ➤➤ ➤➤ CINR versus RSSI (scatter plot and real-time) CINR versus Modulation Scheme and/or MCS Transmit Power versus Cell ID Throughput versus Transmit Power Scheme (antenna 1 and antenna 2) ➤➤ UL ➤➤ UL ➤➤ Modulation 9 September 18.Volume 3 (Detailed Performance Analysis) In our third and final installment we delve much deeper into the KPIs that we captured with the Accuver suite of drive test tools. 16QAM. Further. As we have witnessed in the past there are discernible differences in how each vendor implements a technology. % > 10. and combined) CQI Values versus 64QAM Availability scatter plot ➤➤ MAC-HS ➤➤ MAC-HS ➤➤ MAC-HS ➤➤ HSPA+ ➤➤ MAC-HS Throughput versus DC-HSDPA MAC-HS Throughput (primary. 7.g. Given that we collected network performance data in a number of key markets and that we knowingly included virtually every single vendor + technology combination that exists in North America. Utilization (QPSK. Frankly. the following: ➤➤ Modulation ➤➤ MIMO ➤➤ Volume 3 provides invaluable competitive intelligence while also allowing readers to obtain a greater appreciation for how each technology delivers its results. distribution) – by primary and secondary carriers CQI (average and median) – by primary and secondary carriers Scheduling Success Rate – by primary and secondary carriers ➤➤ HS-PDSCH ➤➤ HS-SCCH ➤➤ Average PHY Layer Served Rate – by primary and secondary carriers PHY Layer Scheduled/Served Rate – DC-HSDPA only ➤➤ Maximum ➤➤ UL Transmit Power (average and median) We will also leverage the capabilities of the XCAP-M tool to analyze these KPIs by several different means. but are not limited to. secondary. Specific topics addressed in Volume 3 may include. subsystem suppliers and mobile operators. potentially including. some vendors do a much better job than their peers. secondary and combined) versus CINR (scatter plot and real-time) versus Cell ID (e. and 64QAM) – by primary and/or secondary carriers as appropriate RI 1 and RI 2 – how MIMO performs at 700MHz Codes (average. Vol. this report provides invaluable competitive intelligence for vendors.
0 36:09. which includes these reports.0 37:21.0 18:09.0 18:01.0 19:29.0 18:41.000 3. $1.0 19:45.495) Full Report – all 3 volumes ($3.0 21:35.RTT TCP dBm dB Mbps ms MOS LTE HSPA+ TD-LTE EV-DO Rev A/B Mobile WiMAX MCS IMO RI BLER 64QAM 16QAM QPSK CQI L1 DL RSCP CINR ACK NACK PDSCH Coming Soon! HS-PDSCH HS-S S E-DPCH FTP UDP HTTP RTT TCP dBm dB Mbp TE HSPA+ DC-HSDPA TD-LTE EV-DO Rev A/B M X MCS MCS SNR MIMO RI BLER 64QAM 16Q CQI L1 DL RSSI RSRQ RSCP CINR ACK NACK PD SCH HS-SCCH MAC-HS E-DPCH FTP UDP HTTP m dB Mbps ms MOS LTE HSPA+ DC-HSDPA TD Rev A/B Mobile WiMAX MCS MCS SNR MIM 4QAM 16QAM QPSK CQI L1 DL RSSI RSRQ ACK NACK PDSCH HS-PDSCH HS-SCCH MA H FTP UDP HTTP RTT TCP dBm dB Mbps ms MO DC-HSDPA TD-LTE EV-DO Rev A/B Mobile Wi CS SNR MIMO RI BLER 64QAM 16QAM QPSK C I RSRQ RSCP CINR ACK NACK PDSCH HS-PD CH MAC-HS E-DPCH FTP UDP HTTP RTT TCP dB ms MOS LTE HSPA+ DC-HSDPA TD-LTE EV-DO obile WiMAX MCS MCS SNR MIMO RI BLER 64Q Pre-order your report license now (included as part of a Signals Ahead subscription) QPSK CQI L1 DL RSSI RSRQ RSCP CINR ACK N HS-PDSCH HS-SCCH MAC-HS E-DPCH FTP RTT TCP dBm dB Mbps ms MOS LTE HSPA+ TD-LTE EV-DO Rev A/B Mobile WiMAX MCS IMO RI BLER 64QAM 16QAM QPSK CQI L1 DL 5.0 19:21.0 19:13.495) Volume 3 – Detailed Performance Analysis ($1.495.300.0 20:17.0 18:17. Number 10 17:29. Vol.000 70% 4.0 21:27.0 20:25.signalsresearch.0 37:33.0 35:33.0 19:53.0 21:20.0 37:45.0 38:09.Cat 8 (1Rx) 20% 10% 0% 18.995) Volume 2 – Quantifying the User Experience ($1.0 37:09.0 36:21.com or write us at Signals Research Group.0 36:57.295.0 Contact Information You may call us at +1 (510) 273-2439 or email us at firstname.lastname@example.org 20:57. You can also visit our website at www.0 . CA 94611 September 18.0 18:33.0 17:45.0 20:49.0 20:41.0 18:57.0 36:33.0 17:37.000 Total DC-HSDPA PHY Layer Throughput (Kbps) 60% 50% 52. $1.9% 35.0 19:37.0 HSPA Demo Phone .295. $3.0 -50 -40 -30 -20 (pre-publishing and post-publishing prices) Volume 1 – Network and Technology Performance ($1. 2011 | Signals Ahead.0 20:09. Subscription information for our Signals Ahead research product.000 0 PHY Layer Primary Throughput (Kbps) PHY Layer Secondary Throughput (Kbps) 36:45.0 18:25.0 20:33.com and we will contact you for your billing information or respond to any further inquiries that you may have. can be found on the last page of this report.995) 10 License Costs 38:21.0 20:01.0 21:05. 7.0 37:57.7% 40% 30% 2.0 21:13.0 17:53.000 1. LLC 10 Ormindale Court Oakland.0 18:49. $1.0 35:57.5% 35:45.0 19:05.
We will cover the more substantive analysis.0 Figure 1. including a 150Mbps server that we provisioned in the immediate Chicago vicinity and a server in Dallas.0 Detailed Results Although we have titled this chapter Detailed Results. our testing took place over a three day period. 21:43. Since we had relatively exclusive access to the pre-commercial network.0 22:31. As previously indicated. and latency.0 21:59. including how AT&T’s LTE network compares with its HSPA+ network and the Verizon Wireless LTE network with a similar test methodology. namely downlink throughput. based on our testing of other commercial LTE networks we believe that it will be quite some time before network loading has any discernible bearing on the performance of the network. 2011 | Signals Ahead. it really focuses only on three KPIs. Number 10 . uplink throughput. we were not concerned about the implications of network loading. 7.0 22:23. That being said.DCS MCS L RSSI SCCH ps ms Mobile QAM DSCH P RTT D-LTE MO RI RSCP AC-HS OS LTE iMAX CQI L1 DSCH Bm dB O Rev QAM NACK UDP DCMCS L RSSI 1 1 3.0 22:07. LLC September 18. “Oh the places we did go!” Geo-plot of test routes with speed (mph) Vehicular Speed (mph) 50 <= x 20 <= x <30 5 <= x <10 40 <= x <50 15 <= x <20 0 <= x <5 30 <= x <40 10 <= x <15 Source: Signals Research Group. in our forthcoming series of reports.0 22:39.0 21:51. including Inter-RAT handover latency.0 22:15. In order to load the pipe. Vol. we tapped into multiple servers across the United States. Texas.
Figure 1A. Come join us! 4G World. Spain Feb. Dallas. Most of the results that we present in this report are based on the Physical Layer throughput. “Oh the places we did go!” (Downtown Houston) Geo-plot of test routes with speed (mph) Vehicular Speed (mph) 50 <= x 20 <= x <30 5 <= x <10 40 <= x <50 15 <= x <20 0 <= x <5 30 <= x <40 10 <= x <15 Source: Signals Research Group. Orange County. with the drive back to the airport (the long pinkish line) covering roughly 20 miles outside of the grid. NV Jan. 20-23 1 2 September 18. LLC In total. Number 10 . Average PHY Layer uplink throughput = 16. To put things into perspective. 10 Invited Speaker Consumer Electronics Show. Barcelona. Including testing that we did on other networks. 8-9 Invited speaker RCR Wireless OC Event. CA Nov. 2011 | Signals Ahead. we covered 205. a good rule of thumb is that the Application Layer throughput using the FTP protocol is 10% lower than the Physical Layer throughput. Vol. Chicago. 10-13 Mobile World Congress.6GB of transferred data. while the long pinkish line extending out from Houston reflects the freeway heading back to Bush International Airport while we were testing the uplink (3. including the results from tests in other markets. Las Vegas. In looking at all of our data. TX Nov. Figure 1 shows the routes that we covered during our downlink and uplink testing. 24-27 LTE Americas. Throughput at the Application Layer will be modestly lower and based solely on the efficiency of the application/protocol that is being used. IL Oct.6 miles while testing AT&T’s LTE network in the greater Houston area.9Mbps). The second figure enhances the testing that was done in the downtown area of Houston. 7. we covered nearly 350 miles. the main focus of our testing spanned a 15 square mile grid.
Vol. All of the testing took place from a vehicle that was constantly moving unless we were stuck in traffic or sitting at a traffic light.6% 60% 20% 0% 2.1Mbps.3.5Mbps 3. Figure 2. LLC Total Distance Traveled = 153.8% 13 September 18.1% 15 . AT&T Houston LTE Network Downlink PHY Layer Throughput Results – CDF and Pie Chart Distribution Average PHY Layer Data Rate = 23.4% 45Mbps 6. Most importantly.6Mbps. Although it is not specifically shown in the figure.1 Downlink Results The downlink results that we present are based on transferring 73.40Mbps 5. We note that the theoretical maximum data rate of a Category 3 device in a 10MHz channel with 2x2 MIMO and under pristine and completely unrealistic network conditions is 73.1Mbps CDF 100% > 50Mbps 4.9GB 0 .6Mbps and the maximum data rate was 61.6Mbps Max PHY Layer Data Rate = 61.5 .2% 80% 45 50Mbps 40 .3% 10 .9% 25 .6% of the time and 21Mbps – the theoretical peak data rate of the operator’s HSPA+ network – for 38.30Mbps 9.9 miles Total Data Transfer = 73.7% 510Mbps 7. Figure 2 shows the results across all test scenarios.5Mbps 1. 7.2. and residential areas in seemingly remote portions of the city.25Mbps 12. The average Physical Layer throughput was 23.4% 5 10 15 20 25 30 Mbps Source: Signals Research Group.3% 2. The average Physical Layer downlink throughput was 23. we believe that we hit all of the major universities and traveled by Reliant stadium on at least a couple of occasions.20Mbps 22. Based purely on luck.5 35 40 45 50 55 60 20 . Number 10 .6Mbps and the maximum data rate was 61. In addition to testing in the greater downtown area. during four tests the average Physical Layer throughput exceeded 30Mbps – we downloaded at least 3GB during each of these tests while driving.15Mbps 14. 2011 | Signals Ahead. areas that can best be described as urban sprawl. the data rate was greater than 5Mbps for 95% of the time.35Mbps 7. we included freeway testing.1Mbps while we achieved at least 5Mbps for 95% of the time.2% of the time.4.9% 40% 30 .9GB of data.4% 35 . The data rate also exceeded 40Mbp for 8.
The second figure shows an enhanced view of the downtown area of Houston. Number 10 .5 35 <= x <40 Source: Signals Research Group. Figure 3.5 <= x <5 40 <= x <45 15 <= x <20 0 <= x <2. Vol. 2011 | Signals Ahead. LLC 14 September 18.Figure 3 contains a Geo-plot of the downlink throughput. 7. AT&T Houston LTE Network Vehicular Mode – Geo-plot of Physical Layer Downlink Data Rates Downlink PHY Layer Throughput (Mbps) 55 <= x 30 <= x <35 10 <= x <15 50 <= x <55 25 <= x <30 5 <= x <10 45 <= x <50 20 <= x <25 2.
5Mbps 11. Vol.5 5 7.5% 1 2. Number 10 .5 10Mbps 6.2 mile drive back to the airport (average vehicular speed = 53mph). Figure 4 contains the results from our uplink testing and Figure 5 contains a geo-plot of the same results.2Mbps Max PHY Layer Data Rate = 23. the average uplink throughput was 16.5% 2.5 .15Mbps 16. 2011 | Signals Ahead.2% 80% 20 .12.2Mbps and the peak data rate was 23.5 10 Mbps 12.9GB of data while driving 51.7 miles.7GB during this test. 7.5 . AT&T Houston LTE Network Uplink PHY Layer Throughput Results – CDF and Pie Chart Distribution Average PHY Layer Data Rate = 15. during the 26.8Mbps.6Mbps. LLC 15 September 220.127.116.11% 10 .2. the second figure expands the results from the downtown area of Houston.5Mbps 3.2 Uplink Results The uplink throughput results are based on transferring 14. Like the downlink results.5 Mbps 0.25Mbps 11.9Mbps – we transferred 3. Sixty percent (60%) of the time the uplink Physical Layer data rate exceeded 15Mbps and 98.7 miles Total Data Transfer = 14. Figure 4.5 15 20 25 0% Source: Signals Research Group.5% 7.6Mbps CDF 100% 1 .5Mbps 1.6% 5 .20Mbps 48. In fact. Our test methodology was identical to our downlink testing. The maximum theoretical Physical Layer uplink data rate for a Category 3 device in a 10MHz channel is 26.9GB 60% 20% 15 .2% of the time it exceeded 5Mbps. the uplink throughput is relatively impervious to vehicular speed.9% 40% 12. The average Physical Layer uplink data rate was 15.5% Total Distance Traveled = 51. Once again.
7.5 2.5 <= x <10 5 <= x <7. Number 10 .Figure 5. Vol. AT&T Houston LTE Network Vehicular Mode – Geo-plot of Physical Layer Uplink Data Rates Uplink PHY Layer Throughput (Mbps) 20 <= x <25 15 <= x <20 12. 2011 | Signals Ahead.5 7.5 0 <= x <=1 Source: Signals Research Group.5 <= x <=5 1 <= x <=2. LLC 16 September 18.5 <= x <15 10 <= x <12.
5 seconds of interrupted data flow) 30 AT&T LTE MAC Layer Throughput (Mbps) 20 AT&T LTE HSPA+ MAC Layer Throughput (Mbps) 10 0 0 50 100 150 200 250 300 Time (sec) 350 400 450 500 550 600 Source: Signals Research Group. 7. Figure 6 shows the MAC-layer throughput for the device when it was on the LTE network and after it switched to the HSPA+ network. Although two tests may not be statistically significant we did observe very similar results between the two handover events – one occurred during downlink testing and one occurred during uplink testing. 2011 | Signals Ahead. It actually took the device only 2. CELL_DCH is the radio resource control (RRC) state used by an HSPA+ device when it is sending large amounts of data.3 Inter-RAT Handover Results We had originally intended to test Inter-RAT handovers between LTE and HSPA+ on our way back to the airport. According to our analysis of the log files.4 seconds to establish a connection to the HSPA+ network. This latter calculation was based on the time between the device signaling that it was leaving the LTE network until the time that it entered the CELL_DCH state.8Mbps AT&T HSPA+ Median Throughput (MAC Layer) = 5. there was ~6. during the process of reviewing the data we stumbled upon two instances where our device handed off from the LTE network to the HSPA+ network.3.9Mbps Mbps 50 40 LTE to HSPA+ CELL_DCH Handover Time = ~2.4 seconds (~6. We suspect. We used the MAC layer throughput since it provided greater resolution and since it was a good common denominator between the two air interfaces. Number 10 .5 seconds of down time in the data transfer as the device was switching network. Figure 6. but as it turned out the operator had coverage the entire way. that the coverage primarily existed along the freeway and not in the residential areas extending out from the freeway. but didn’t confirm. LLC 17 September 18. Vol. Inter-RAT Handover AT&T LTE Median Throughput (MAC Layer) = 16. However.
we have been disappointed with the latency on the LTE networks in North America. This requirement can impact the latency. such as video streaming.50ms 8% 45 .55ms 17% 48 . Latency is an area for improvement. That being said. Without naming names. we also measured latency to these servers using a local broadband service provider. As indicated in the two pie charts.47ms 8% 50ms 6% Source: Signals Research Group.65ms 4% 8% AT&T LTE Latency Stationary Test) Average Latency = 49ms >65ms 61 . In our forthcoming reports we also encourage readers to focus on the relative results versus the absolute results. we used Verizon FiOS from a residence in Dallas to test the server that we used for latency testing in Houston and Dallas.45 . Figure 7 shows results from two latency tests that we did on AT&T’s LTE network in Houston. but this result is not reflected in the below pie charts. generate relatively large buffers which can endure a few seconds of interrupted data transfers. At the moment. One test was done from a fixed location while the other test was done from a moving vehicle. 3. albeit when we were testing a commercial 3G network with a large number of subscribers.From our perspective. the handover latency was quite good.47ms 15% 43 . and AT&T isn’t alone in that regard. in particular given the early days of the network and the infancy of LTE. In order to be as objective as possible. the average latency was 49-50ms over an extended period of time (>60 seconds) with 40ms being the minimum latency.60ms 21% 51 . Vol. This opinion is based on our experience with LTE networks in Europe (15ms average latency to an external server) and the results from some of the 3G networks that we have been testing in North America.4 Network Latency Results Our methodology for testing network latency leveraged local servers in the markets that we were testing. Inter-RAT handover latency will have to get much better in order to deliver a seamless voice service between the two networks.45ms 22% 40 . Figure 7.60ms 23% 43 .42ms 17% 56 . 2011 | Signals Ahead. in particular if the peering points between the operator’s mobile network and the Internet are less than optimal. Since VoLTE on AT&T’s network is still several years away there should be ample time to address the issue. 7. we documented a minimum round trip latency of 27ms while testing in Houston using this very same server. LLC 18 September 18. At one point we saw a low of 35ms latency. For example. handovers from HSPA+ to LTE require that the device return to the IDLE state. On occasion we tried multiple servers in order to obtain a mix of results.6% 65ms 6% 56 . AT&T Houston LTE Network Latency Results – Pie Chart Distribution AT&T LTE Latency (Drive Test) Average Latency = 50ms >65ms 61 . In order to test all networks simultaneously we also had to use servers that resided outside of an operator’s network.55ms 9% 48 . In general. A typical user would never notice the interruption with most applications – network drive testing being a notable exception.42ms 18% 40 . Number 10 .45ms 18% 51 . Even applications.
Doing the math.9 kilobytes of data.4MB of data [we did the test and examined the Accuver log files to reach this conclusion].800 people in the Houston metropolitan area.5 The Statistical Significance of 90GB In total. suggesting that the actual number of implied tests is considerably higher.912 web pages. to a somewhat recent study by Google. a typical web page consists of 320kB of data. 19 September 18.750 bandwidth tests using this service. Presumably. 2011 | Signals Ahead. suggesting that we loaded the equivalent of 294. including the user experience tests that we conducted – the results from these tests are not included in this report.33GB of data.946. Vol. meaning that our data usage was the equivalent of each Houstonian transferring 15. ➤➤ According ➤➤ We transferred 5GB for each gallon of gas that our rental SUV consumed. but consider the following: ➤➤ 90GB ➤➤ There ➤➤ A equates to nine months of usage without incurring overage fees. This last metric may say more about the popularity of Katy Perry than the performance of AT&T’s LTE network. ➤➤ For each and every Katy Perry song that we listened to on the local radio station (27 times) we transferred 3. we transferred 90 GB during our testing of AT&T’s LTE network. we did the equivalent of 1. Readers may or may not agree that 90GB of transferred data can result in statistically significant results. 7. the amount of transferred data is a function of the capabilities of the network. are 5. Number 10 . typical downlink/uplink/latency test using the Speedtest.3. This metric may say more about the fuel inefficiency of a Ford Explorer than the efficiency of LTE.net portal when the network being tested supports a 50/15Mbps pipe uses 52.
We used the Accuver XCAP post-processing tool to analyze the data and to help us create the figures which appear in this summary report. respectively. 7. Number 10 . augmented by the Accuver XCAL drive test tool to collect the underlying performance indicators and to conduct the user experience tests. Figure 8. 2011 | Signals Ahead. We used the Accuver XCAL-MO and XCAL-M tools to collect the underlying performance indicators and the Accuver XCAP-M post-processing tool to do the analysis of the data that we collected. Vol. We have included two figures since they also help prove that we observed downlink data rates greater than 61Mbps (Figure 7) and uplink data rates in excess of 23Mbps (Figure 8). Thanks to a combination of the powerful tool and countless hours spent on the road. or 55mph and 64mph. Figure 7 and Figure 8 illustrate a typical user display that we used when collecting the data.0 Test Methodology For the drive tests that we have been conducting this summer we primarily used the recently released Accuver XCAL-MO network benchmarking tool. For purposes of our tests. we were traveling in excess of 60mph when we achieved both results. In theory we could have installed multiple dongles for each network/technology. Interestingly.4. we are convinced that we have witnessed network performance – both good and bad – that would have otherwise not been observed. we “limited” the XCAL-MO to only four dongles – one dongle for each network/technology that we wanted to test. XCAL-M Drive Test Tool in Action – DL performance Source: Accuver XCAL and SRG 20 September 18.
we established multiple sessions in order to offset some of effects of transport latency and the TCP ACK window associated with FTP. Vol. including servers in Phoenix (>100Mbps). Testing in each market took place from as early as 4AM local time until the early evening hours. meaning that the super-high data rates associated with LTE. Since we were using test equipment we had the ability to determine whether or not network loading 21 September 18. For our “Mother of all Network Benchmark Tests” each operator provided us with at least two dongles. network latency tests were done to servers located in the vicinity of the market in which we were doing the tests. For all of our testing. As we have documented in the past. even with the legacy operating system. we received multiple dongles. we used the Windows 7 operating system. are not always achieved. we believe Windows 7 leaves something on the table. Number 10 . Since we can’t rule out the effect of transport latency. such as Clearwire. meaning that the data rates could already be limited. Testing in each market took place from as early as 4AM local time until the early evening hours. 2011 | Signals Ahead. with multiple network/technology deployments (e. At our request. We also note that the connection manager associated with at least one operator’s technology/network does not change the TCP window size when it is installed. XCAL-M Drive Test Tool in Action – UL performance Source: Accuver XCAL and SRG We used the Windows 7 operating system. 2x20MHz LTE and 1x10MHz Mobile WiMAX). although in the case of operators. in particular with a 20MHz channel.Figure 9. which uses a dynamic window allocation methodology to theoretically deliver the best possible throughput for the given combination of latency and channel conditions. 7. For purposes of these tests. In order to ensure that we fully loaded the air link channel for the networks/technologies that we were testing. readers should focus their attention on the relative performance differences versus the absolute results. which uses a dynamic window allocation methodology to theoretically deliver the best possible throughput for the given combination of latency and channel conditions.. We believe our approach is still appropriate given that it is logistically impossible to purchase a notebook computer with the Windows XP operating system. We also conducted user experience tests from the confines of our hotel room in the dead of night.g. Dallas (300Mbps) and Chicago (150-300Mbps). AT&T provided us with an LTE dongle and unlimited access to its LTE network. Further. we leveraged multiple high-bandwidth servers.
ABS and the importance of interference cancellation in the handset. but in the interim CSFB isn’t even working as promised. we offer an in-depth look at the future of HSPA+ (Release 11 and beyond). including TD-LTE. We compare DC-HSDPA with HSPA+ performance in a number of side-by-side tests. tion with Sanjole we examine how some of today’s commercial LTE eNodeBs allocate network resources when serving multiple devices. including the use of cloud computing.0. We also get into the technical details of how HetNet works.In Case You Missed It ➤ 7/6/11 “Mobile Platforms – the center of mobile networks” In this report we discuss the recent trends impacting the various mobile platforms that exist and what has transpired since our piece from three years ago on Web 2. MediaTek and HiSilicon. who provided us with its XCAL-W drive test tool and XCAP-W post-processing software. collide” In addition to covering the basics of heterogeneous networks (HetNet). or the lack thereof. including discussions on eICIC. 2011 | Signals Ahead. 7. We also discuss the growing trend of operators who are intelligently adding more capacity to their networks through the use of higher performing devices/chipsets and upgrades to their network infrastructure. fragmented we fail” We ➤ 1/12/2011 DC-HSDPA: “Double the Bandwidth. but the emphasis seemed to be on how to best move away from the technology and adopt TD-LTE. Double ➤ 12/10/2010“Can you schedule me now?” In collabora- ➤ 5/16/2011 “HetNet: When big cells and small cells ➤ 12/3/2010 “A Perspective from LTE Americas and the ➤ 4/26/2011 “Chips and Salsa XIII: Now Seasoned with ➤ 11/4/2010 “A G-Wiz World” We provide drive test results ➤ 3/15/2011 “Looking beyond HSPA+: keeping up with ➤ 10/7/2010 “2x20MHz of LTE and the Heisenberg 22 September 18. thus blurring the performance differences between the two technologies. Uncertainty Principle” We provide an update on LTE network performance based on extensive testing of TeliaSonera’s LTE networks in Stockholm and Gothenburg. is still on everyone’s minds. starting as early as 2015. Double the Pleasure. provide the key takeaways from the LTE World Summit. Latency and the impact of new features on legacy devices are two areas of prime importance where HSPA+ could face challenges relative to LTE. representing chipsets from 9 different suppliers. ➤ 1/12/2011 “DC-HSDPA: Double the Bandwidth. Part I” In collaboration with Accuver. In the most recent benchmark study we tested 16 different device configurations. Number 10 . VoLTE. and that further improvements are necessary if LTE is going to support multiple users in an efficient manner. such as Samsung (HSPA+). Ultimately. we look at what is being done with legacy 3G femtocells to limit interference-related problems that they introduce. we conclude that many of the features that are being incorporated into LTE will find their way into HSPA+. based on a total of 42 HSPA+ test scenarios and 26 HSPA test scenarios. We address the state of the mobile platforms that exist. Soy Sauce” In collaboration with Spirent Communications we provide results from the industry’s only independent performance benchmark tests of HSPA+/HSPA chipsets. the Pleasure. although a growing focus on the use of 1800MHz for those operators that have access to it is encouraging. discuss the debate about a smart or dumb pipe strategy. we provide results and analysis from an extensive drive test of Telstra’s DC-HSDPA network. who provided us with its XCAL-W drive test tool and XCAP-W post-processing software. We provide the results. ➤ 6/8/2011 “United we stand. the Joneses” Based on interviews with 3GPP member companies and a thorough review of 3GPP submissions. We provide an LTE market update. network offloading and data caching. Finally. and the impact of smartphones and social networking services. Finally. there was a lot of talk about Mobile WiMAX. Part II” In collaboration with Accuver. including new entrants. held in Amsterdam. We determine that while LTE may deliver a compelling user experience. we present a compelling series of analytical studies which demonstrate the need for macro network offload. Vol. The 60+ page report provides detailed results and analysis based on more than 600GB of transferred data. we provide results and analysis from an extensive drive test of Telstra’s DC-HSDPA network. both with the macro network and between each other. it is largely due to an empty network and the large channel bandwidths. GSMA Mobile Asia Congress” We provide and discuss various data points which stem from our participation at the LTE Americas event in Dallas and the GSMA MAC event in Hong Kong. provide our thoughts on the current and future prospects and look at the various trends that are driving the industry. a key LTE-Advanced (R10) feature. We compare DC-HSDPA with HSPA+ performance in a number of side-by-side tests. Spectrum fragmentation tops the list of key LTE topics. Sweden. for TeliaSonera’s HSPA+ network in Sweden and provide data points from this year’s 4G World event in Chicago. intelligent networks. Intel (HSPA+).
Further. A large percentage of our test data was collected from a moving vehicle. this latter approach achieves anecdotal results which are statistically meaningless and not necessarily reflective of the overall network performance. Although we do not include results from our user experience tests in this summary report. 7. accessing the broadband wireless network from a moving vehicle will be more commonplace now than in the past. moving a few feet or turning 90 degrees can meaningfully impact the achievable throughput. we received no sponsorship or funding from the participating companies. We also used popular applications. Ideally. We received no sponsorship or funding from the participating companies. By testing from a moving vehicle and by transferring hundreds of Gigabytes of data. Number 10 . SRG takes full responsibility for the analysis and conclusions that are documented in this report and in our forthcoming series of reports. One drawback of our approach is that it does tend to understate the performance of the network since the fading conditions from a moving vehicle are more challenging than from a stationary position or someone walking down the street. HSPA+. but this approach goes well beyond something that we can reasonably do for these studies. Vol. we would include stationary tests from hundreds of locations in a given market and all times of day. and content providers. we based our analysis and conclusions on literally hundreds of Gigabytes of transferred data. was impacting the results. MIMO availability. we used popular websites for our HTTP web page download tests and Google Mail for the email application. we believe that our conclusions are statistically significant. This approach ensured that we achieve statistically meaningful results since as we have demonstrated in past reports. scheduling frequency. Later in the day. in order to maintain our independence. We also could not have done this report without the support of Accuver who provided us with its suite of drive test tools and post-processing software. This sampling of the network performance also provides no insight whatsoever into how/why the throughput was achieved since the KPIs are limited to throughput and latency. even if the results may slightly understate the true capabilities of the network. From our perspective. Mobile WiMAX. As such. 23 September 18.One drawback of our approach is that it does tend to understate the performance of the network. we foot the bill for all of our travel expenses not to mention an inordinate amount of time and effort collecting the data and writing these series of reports. such as Skype. Suffice it to say that in the early morning hours network loading was not a concern for any of the networks. network loading impacted the performance of certain networks/technologies while it was not even a consideration with other networks/technologies. etc. etc. Given the emergence of next-generation smartphones (LTE. such as Netflix and YouTube. versus KPIs. We take this phenomenon into consideration when doing our analysis. such as modulation type. number of assigned resource blocks. This approach is markedly different from the more commonly used method which involves using popular web-based “speed testing” sites and transferring tens of Megabytes of data. 2011 | Signals Ahead.). Like all Signals Ahead reports. to evaluate the impact of various video streaming services.
Previously. Mr. 24 September 18. Mr. Mr. He has been an invited speaker at industry conferences around the world and he is frequently quoted by major news sources and industry newsletters. Bloomberg News. Until next time.5.0 Conclusions Now that we have finally finished our testing we will in short order be releasing the first in a series of reports. including its widely acclaimed “Chips and Salsa” series of reports that focus on the wireless IC industry. Graduate School of Business. Thelander has also published numerous articles for leading trade publications and engineering journals throughout his career. and The China Daily. The latter option is more economically attractive since it includes at least fourteen additional issues. If you are currently a subscriber to Signals Ahead then you will receive these reports as part of your subscription. We expect that the first report will be released before the end of September and that the following two reports will follow in October and potentially extending into November. Investors Business Daily. Mr. Thelander earned a Masters of Science in Solid State Physics from North Carolina State University and a Masters of Business Administration from the University of Chicago. Number 10 . The Wall Street Journal. In his current endeavor he leads a team of industry experts providing technical and operator economics analysis for clients on a global basis. Thelander was a consultant with KPMG (now known as BearingPoint) and a communications officer with the United States Army. 7. Thelander is also responsible for the consultancy’s Signals Ahead research product. Thelander was an analyst with Deutsche Bank Equity Research. If you are not a subscriber then you can purchase the reports on an individual basis or purchase a corporate subscription to our research. be on the lookout for the next Signals Ahead…. Michael Thelander Michael Thelander is the CEO and Founder of Signals Research Group. Vol. 2011 | Signals Ahead. Reuters. Mr. Prior to joining Deutsche Bank. including The Economist.
please fell free to contact Signals Research Group. the Platinum package includes the Global License. 7. This report is provided for informational purposes only and on the condition that it will not form a basis for any investment decision. all rights reserved by Signals Research Group. plus up to five hours of analyst time. CA 94611 Our fax number is (510) 338-1284.495) ❒ Visa ❒ MasterCard Check Number PO Number Title: Phone: ( ) Credit Card # Exp Date / / Mailing Address Signals Research Group. estimates. LLC seeks to do business and may currently be doing business with companies discussed in this report.signalsresearch. LLC has no obligation to update. LLC – ATTN: Sales 10 Ormindale Court Oakland. LLC and are based on our understanding of past and current events shaping the wireless industry. Signals Research Group. Readers should be aware that Signals Research Group. Such investment activity may be inconsistent with the analysis provided in this report.995) ❒ Global License ($7. Terms and Conditions: Any copying. changes or subsequently becomes inaccurate. This report may not be reproduced. We will not process your payment until after the trial subscription period is completed. Other packages are available. The Group License includes up to five users from the same company. LLC. 2011 | Signals Ahead. The information has been obtained from sources believed to be reliable. Vol. copied. Additional information and disclosures can be found at our website at www. including unauthorized sharing of user accounts. Alternatively. projections or forecasts in this report constitute the current judgment of the author(s) as of the date of this report. Signals Research Group. Finally. 25 September 18. or any opinion. or republishing of this material. LLC and/or its affiliates/investors may hold securities positions in the companies discussed in this report and may frequently trade in such positions. Opinions. modify or amend this report or to otherwise notify a reader thereof in the event that any matter stated herein.Signals Ahead Subscription The Signals Ahead newsletter is available on a subscription basis. LLC (copyright ©2011.com. projection.995) Payment Terms ❒ American Express ❒ Check ❒ Purchase Order Name: Affiliation: Mailing Address: ❒ Platinum ($9. LLC makes no representation as to the accuracy or completeness of such information. Reference in the newsletter to a company that is publicly traded is not a recommendation to buy or sell the shares of such company. please note disclaimer : The views expressed in this newsletter reflect those of Signals Research Group. We offer four distinct packages that have been tailored to address the needs of our corporate users. analysis or interpretations of events are inaccurate. you may contact us at (510) 273-2439 or at information@signalsresearch. Corporate Rates (18 issues) ❒ Group License ($3. LLC might have a conflict of interest that could affect the objectivity of this report. The Global License is the most attractive package for companies that have several readers since it is offered to an unlimited number of employees from the same organization. redistributing. We are always seeking a more accurate understanding of the topics that influence the wireless industry. Number 10 . LLC). Signals Research Group.com and we will contact you for your billing information. is strictly prohibited without the written consent of SRG. forecast or estimate set forth herein. If you feel our opinions. but Signals Research Group. distributed or published without the prior written authorization of Signals Research Group.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue listening from where you left off, or restart the preview.