Initial Access in 5G mm-Wave Cellular Networks

arXiv:1602.07731v1 [cs.IT] 24 Feb 2016

Marco Giordani† , Marco Mezzavilla , Michele Zorzi†
† University of Padova, Italy 
NYU Tandon School of Engineering, Brooklyn, NY

Abstract—The massive amounts of bandwidth available at
millimeter-wave frequencies (roughly above 10 GHz) have the
potential to greatly increase the capacity of fifth generation
cellular wireless systems. However, to overcome the high isotropic
pathloss experienced at these frequencies, high directionality
will be required at both the base station and the mobile user
equipment to establish sufficient link budget in wide area networks. This reliance on directionality has important implications
for control layer procedures. Initial access in particular can be
significantly delayed due to the need for the base station and the
user to find the proper alignment for directional transmission
and reception. This paper provides a survey of several recently
proposed techniques for this purpose. A coverage and delay
analysis is performed to compare various techniques including
exhaustive and iterative search, and Context-Information based
algorithms. We show that the best strategy depends on the target
SNR regime, and provide guidelines to characterize the optimal
choice as a function of the system parameters.
Index Terms—5G, initial access, millimeter-wave, cell search,
cellular systems, coverage, discovery delay, context information

I. I NTRODUCTION

T

HE fifth generation (5G) of cellular systems is positioned
to address the user demands and business contexts of
2020 and beyond. In order to face the continuing growth in
subscribers’ demand for a better mobile broadband experience,
there is a need for significant performance improvements, such
as for example [1]: (i) much greater throughput, with data rates
of at least 1 Gbps or higher, to support ultra-high definition
video and virtual reality applications; (ii) much lower latency,
less than 1 ms, to support real time mobile control and Deviceto-Device (D2D) applications; (iii) ultra-high reliability and
much higher connectivity, to provide seamless service everywhere; (iv) lower energy consumption, reduced by a factor of
1000, to improve the battery life of connected devices.
In order to deal with these requirements, some key aspects
have been identified to make this future network a reality.
Since current micro-wave (µW) spectrum under 5 GHz is
fragmented and crowded, there has been significant interest in
the millimeter-wave (mm-Wave) bands above 10 GHz, where
a vast amount of largely unused spectrum is available. On one
hand, the enormous amount of available spectrum can support
the higher data rates required in future mobile broadband
access networks. Moreover, the physical size of antennas at
mm-Wave frequencies is so small that it becomes practical
to build very large antenna arrays (e.g., ≥ 32 elements) to
provide further gains from spatial isolation and multiplexing.
On the other hand, the increased carrier frequency makes the
propagation conditions more demanding than at the lower frequencies traditionally used for wireless services. For example,

blockage becomes an important issue to take care of, as mmWave signals, due to their small wavelengths, do not penetrate
most solid materials very well (e.g., buildings made of brick)
and are subject to very high signal attenuation [2]. Another
pillar of 5G will be to use many more base stations, deployed
according to a heterogeneous network (HetNet) paradigm,
combining macro sites with smaller base stations and using
a wide range of radio technologies. These will include LTE,
Wi-Fi and any future 5G technologies, integrated flexibly in
any combination.
In this context, the definition of new control layer procedures is critical, in particular initial access (IA), which
allows a mobile user equipment (UE) to establish a physical link connection with a base station (BS), a necessary
step to access the network. In current LTE systems, IA is
performed on omnidirectional channels, whereas beamforming
or other directional transmissions can only be performed after
a physical link is established. On the other hand, in order
to overcome the increased isotropic pathloss experienced at
higher frequencies, in 5G mm-Wave cellular systems the IA
procedure must provide a mechanism by which the BS and the
UE can determine suitable initial directions of transmission.
However, directionality can significantly delay the cell search
and access procedures, which is a particularly sensitive issue
in 5G networks, and thus motivated us to identify and study
some performance trade-offs, in terms of both delay, coverage
and overhead.
This work provides a survey of recent directional IA techniques for mm-Wave cellular systems. As an extension of
our previous work [3], in the present paper we compare various search schemes, including exhaustive search, an iterative
scheme that successively narrows the search beamwidth, and
Context-Information (CI) based algorithms, where users are
informed about the surrounding mm-Wave BSs geolocations
through a µW link. We compare the performance of these
approaches in terms of both misdetection probability and
discovery time, under some overhead constraints and as a
function of the channel conditions. Our results show that
the optimal strategy depends on the target SNR regime and
provide some guidance about the best scheme to use, according
to each scenario.
The paper is organized as follows. In Section II, we list some
of the crucial reasons why the current LTE IA procedure is
not suitable for future mm-Wave systems. In Section III, we
review some of the most important contributions related to IA
in mm-Wave cellular systems, while in Section IV we describe
in detail the three initial access procedures we are going

b) Multi-connectivity: To ensure sufficient coverage. considering different scanning and signaling procedures. which may be related to obstacles. Additionally. A natural (and practical) solution is to use beamforming even in the first stages of the initial access procedure. where base stations periodically transmit synchronization signals. Hence. in order to enable mmWave cells to directly steer towards the user position. we will focus on the cell search phase. In this study we focus on analog beamforming techniques (considered to be more energy-efficient than their digital . The anchor BS gets control over IA informing the booster BS about user locations. in order to alleviate the exhaustive search delay issue. In this section we list the main factors that make 4G-LTE procedures unsuitable for use in a 5G mm-Wave context. This issue already emerged in HetNets. to evaluate access delay and system overhead. since omnidirectional signaling may generate a mismatch between the (relatively short) range at which a cell can be detected in this case. R ELATED WORK Papers on IA in mm-Wave 5G cellular systems are very recent. as will become clear in the following. we presented a comparison between the exhaustive and the iterative techniques. we expand the analysis to a CI-based algorithm and describe a proposed enhancement. In our study. a directional cell discovery procedure is proposed. when considering an analog multiantenna architecture. 4G-LTE: INITIAL ACCESS LIMITATIONS In all mobile communication systems. are not suitable approaches in mm-Wave networks. the challenge with higher frequencies is the need to also account for dynamics such as directionality and intermittency. which are provided by a separate control plane.11ad WLAN and WPAN scenarios. in order to improve the cell discovery procedure and minimize the delay [11]. where a faster user discovery technique is implemented. the analysis demonstrates significant benefits of low-resolution fully digital architectures in comparison to single stream analog beamforming. II. In this work. In [5]. multi-path scenario. when considering a realistic dense. In [12]. as introduced in [7]. and adapt them to the upcoming mm-Wave related challenges in order to overcome such limitations. it may instead be essential to exploit the beamforming gains even during the CS procedure. Consequently. Finally. for example. new adaptive techniques have to be introduced. d) Dynamics-aware access: Due to denser topologies. IV. in [13]. or waiting for a random back-off time (as done in LTE). dominates the overall delay performance. In [3]. since research in this field is just in its infancy. Most literature refers to challenges that have been analyzed in the past at lower frequencies in ad hoc wireless network scenarios or. as signals are transmitted omnidirectionally in the downlink and beamforming is used only after a physical link has been established. where the authors proposed an exhaustive method to sequentially scan the 360◦ angular space. Such a comparison is the main goal of this paper. hand rotations. there is an urge to extend current LTE procedures. acquiring time-frequency domain synchronization during CS is facilitated. which determines whether or not a UE is able to detect a BS and. On the other hand. we evaluate three IA procedures. The technical issues described in this section call for new initial access procedures and for a detailed assessment of their performance in realistic 5G mm-Wave scenarios. In Section V we evaluate through simulations some comparison metrics. On the other hand. more recently. Increasing the transmission power. whereas blockage causes a failed message delivery due to a channel drop. and other mm-Wave-sensitive events. extraction of system information. Initial access design options are also compared in [9]. urban. to discriminate among different failed-access causes. in terms of both misdetection probability and discovery delay. However. or come up with new algorithms and new methods. including at least a macro BS operating in the µW spectrum.to compare. mm-Wave networks will be much denser. in Section VI we summarize our major findings. and random access (RA) [4]. such as misdetection probability and discovery time. to scan the angular space. Furthermore. thereby losing the broadcast propriety of the wireless medium. an evolution of [11] is presented. in [8]. paper [10] presents a two-phase hierarchical procedure. and the (much longer) range at which a user could directionally send and receive data [5]. The initial access problem in mm-Wave cellular networks has been considered. Deafness refers to a situation where the transmit-receive beams do not point to each other. Hence. III. conventional reference signal received power (RSRP)-based association schemes would be highly inefficient in mm-Wave cellular networks. potentially in time-varying random directions. the IA procedures have to be redesigned in order to capture this fundamental new feature. a terminal transitioning from IDLE to CONNECTED mode must perform the following steps: cell search (CS). under overhead constraints. [6]. IA messages may not be received due to deafness or blockage phenomena. Our goal is to compare multiple IA procedures under an overhead constraint and to derive the best trade-offs. in the 60 GHz IEEE 802. directionality means that only one direction can be considered at a time. keeping in mind that a fully directional data plane requires a directional IA procedure in the new frequency band. Each user is expected to simultaneously detect multiple potential serving stations. showing how to capture the effects of position inaccuracy and obstacles. with important implications for protocol design and delay performance that must be carefully taken into consideration. In mm-Wave bands. Context-Information based procedures aim at exploiting knowledge about user and/or BS positions. a) Discovery range mismatch: In LTE systems. c) Deafness and blockage: In mm-Wave cellular networks. I NITIAL ACCESS IN 5G MM -WAVE NETWORKS As summarized in Table I. booster cells (operating at mm-Waves) are deployed under the coverage of an anchor cell (operating at microwaves).

2) Each UE gets its own GPS coordinates (this will require a certain energy cost). We adopt analog beamforming. counterparts [6]. 3) According to the information obtained in Steps 1 and 2. Context-Information based search This algorithm is articulated into three main stages [11]. • • • Low discovery delay. Moreover. which allows steering towards one direction at a time. The evaluation of hybrid and fully digital beamforming architectures and a comparison among them are left for future work. The goal is to identify the best TX-RX beam for each BS-UE pair. . [10] and [3] for further procedural details on exhaustive and iterative techniques. we will assume a static deployment where no handover management or UE motion tracking is required. able to detect the intended signal in more noisy channels. In order to detect a PSS signal. P ERFORMANCE EVALUATION To conduct our performance analysis. This is a more natural way of using CI and assigns the burden of beamscanning to the BS (which would have to do it anyway in the presence of multiple users) rather than to the UE (which can in this case save energy). number of spatial clusters. Table I: Summary of the three IA algorithms compared in this work. • • • Pure CI-based search Low discovery delay. Statistical models are derived for key channel parameters including pathloss. the BS transmits pilots over wider sectors. Each signal has a minimum duration Tsig = 10µs. Not suitable for edge users. taken to be −5 dB in our results. The primary synchronization signal (PSS) is transmitted periodically. respectively. Decreasing τ would allow to find more users. The channel model is based on recent real-world measurements at 28 GHz in New York City [14] to provide a realistic assessment of mm-Wave micro and picocellular networks in a dense urban deployment. for a duration of Tsig seconds. and to [11] for CI-based 1 We note that the procedure we consider in this paper is different from [11] in that the CI is available at the UE and concerns the location of the BS. once every Tper seconds. Bad coverage. we assume a slot structure similar to the one described in [9]. and misdetection probability (PMD). We refer to [8]. we evaluate the performance in terms of discovery delay. where each transceiver can look at only one direction per slot. and a noise figure of 5 dB is assumed. Exhaustive search A brute-force sequential beam searching [8]: both users and base stations have a predefined codebook of N directions (each identified by a BF vector) that cover the whole angular space and are used sequentially to transmit/receive. which is the time required by a BS to discover a UE in its coverage range. implemented through a Uniform Planar Array (UPA). Suitable for edge users. High discovery delay. 1 1) The macro BS (at µW frequencies) spreads omnidirectionally the GPS coordinates of all the mm-Wave stations within its range. Meanwhile. while in the second phase it refines its search within the best such sector by steering narrower beams. The total system bandwidth is taken to be 1 GHz. A more detailed discussion and comparison between the two paradigms is beyond the scope of this paper. Only LOS scenarios. which is deemed to be sufficient to allow proper channel estimation at the receiver. the transmission powers of BS and UE are set to 30 dBm and 23 dBm. Iterative search A two-stage scanning of the angular space [10]: in the first phase. each UE geometrically selects the closest BS to connect to and steers a beam towards the direct path. at the cost of designing more complex (and expensive) receiving schemes. angular dispersion and outage.Exhaustive Search Iterative Search • • • Good coverage. each mm-Wave BS performs an exhaustive search to detect the best transmit-receive direction. Further details of this model and its parameters can be found in [15]. V. which is the probability that a UE within the cell is not detected by the BS. In this study. the SNR received at the UE has to lie above a certain threshold τ . Cost of getting GPS coordinated. algorithm.

Tsig Total delay Ns · Tsig /φov User at 95 m Exh. the best technique generally depends on the target SNR and on the considered scenario. where BS always use narrower beams that provide higher gains.g.UE Figure 1: PMD for exhaustive and iterative techniques. SNR threshold τ = −5 dB. we observe that. 64 × 4 80 400 µs 640 ms Exh. so that UEs belonging to that sector can accumulate more energy.6 ms It. we compare the CI-based initial access technique with the two sequential schemes. 0. whereas iterative search may be preferred otherwise (e. the BS transmits its PSS for a longer time in the same sector. it may not be desirable to implement exhaustive procedures. within the 0 ÷ 30 m range. where links are often NLOS. < 100 150 200 Distance BS . However. Nonetheless. in bigger cells. Conversely. Procedure BS antennas UE antennas Ns Delay Exh. In Subsection V-B. when considering smaller cells.6 ms It. The key finding is that pure CI algorithms may not be suitable for urban scenarios. and is computed as φov = Tsig /Tper . 64 × 16 4 in 1st phase 64 in 2nd phase 16 44 8. and is reported in Table II. as in the case of cell-edge users). The total delay in Table III captures both the required number of slots and the PMD specifications. the corresponding Tper = Tsig /φov is selected in order to have a constant overhead φov = 5 percent. -1 10 Iterative 64×4 Iterative 64×16 Exhaustive 64×4 Exhaustive 64×16 10-2 10-3 10-4 0 Misdetection probability: Since BSs in iterative search transmit over wider beams in the first phase. As shown in Figure 1.8 ms Exh. From these results we can derive the minimum signal length to meet a certain PMD requirement by reading the abscissa of the intersection of each curve with the horizontal line. which bring higher discovery delays.01 for users at 95 and 35 meters from the BS. the misdetection probability is fairly small.g.8 ms 1st It. 64 × 4 28 > 3160 µs > 1760 ms It. to guarantee PMD < 0. 64 × 16 144 10 µs 28. vs. 64 × 16 44 50 µs 44 ms Table III: Total delay. 64 × 4 80 13 µs 20. Trade-off between delay and PMD: In order to guarantee the misdetection probability below a certain threshold (i. Hence. when UE receives in 4 or 8 directions. Sequential approach Discovery delay: We consider a minimum signal duration Tsig = 10 µs and a target overhead of φov = 5 percent. in case of smaller cells). 64 × 16 44 1580 µs 1390 ms User at 35 m Exh. the BS-UE distance. we see that the corresponding curve intercepts the PMD . The main conclusion of this study is that exhaustive search is likely to be the best IA configuration. we compare two sequential-based IA schemes: the exhaustive search and the iterative technique. the iterative approach requires fewer slots because BSs do not use narrow beams to scan the whole 360◦ angular space. Given that an IA technique requires Ns slots to be sent (one per Tper ). considering a UE at a distance of 95 and 35 meters from the BS. According to the specific Tsig values obtained. As expected. we can increase the signal duration: if Tsig is increased. UEs far from the BS will experience high misdetection probabilities...8 ms It. which results in a reduced beamforming gain. the misdetection probability is higher if compared to exhaustive search.e. resulting in higher SNR and correspondingly reduced PMD. 50 Ns min. 64 × 16 64 16 144 28. especially if we want to provide good coverage with high probability at relatively large distances (e. Procedure Table II: Discovery delay for different sequential-based IA techniques.A. we show that more sophisticated directional procedures have the potential to reduce the total discovery delay and grant good coverage. 64 × 4 4 in phase 64 in 2nd phase 4 28 5. For example. respectively. 64 × 4 64 4 80 16 ms Exh. for the exhaustive 64 × 16 case in Figure 2(a). motivating further efforts towards designing IA methods to grant access to celledge users without excessive delay. 64 × 4 28 160 µs 89. 64 × 16 144 125 µs 360 ms It.8 ms 0 10 Misdetection probability In Subsection V-A. The two plots in Figure 2 show the PMD as a function of Tsig for the different schemes..01). the delay can be computed as Ns · Tsig /φov . The overhead refers to the percentage of time the channel is used to send digital information across the functional interface between the UE and the BS for the purpose of controlling the IA procedure.

to meet the misdetection requirements. a slot of duration Tsig ' 150 µs is sufficient to meet the PMD requirements. and for this reason require a longer signal duration. in this case the CI-based algorithm may require very long signals to get enough energy. exhaustive searches can operate with shorter slots but require more of them. which was the better approach according to the results in Table II. signal duration Tsig . and leave a study in the presence of CI errors for future study. Therefore. On the other hand. vs.01) for users at 95 m from the BS.01) Misdetection probability Misdetection probability 10 -1 10 10-2 -3 10 3160 316 100 31 10 Iterative 64×4 Iterative 64×16 Exhaustive 64×4 Exhaustive 64×16 PMD threshold ( = 0. the beam chosen by the UE may actually be suboptimal. However. the total discovery delay can be evaluated as follows: 1) For a pure CI-based search. Conversely. we assume that Context-Information is not affected by any GPS error. To meet the PMD requirements.2 In general. if the direct path does 2 We remark that these computations can also be applied to evaluate the energy consumption performance. a signal duration of around 1580 µs must be used. resulting in a delay equal to Ns · Tsig /φov ' 192 ms. CI-based initial access We first consider the pure CI-based technique reported in Table I. Moreover. when considering the iterative 64 × 16 case.01) -1 10 -2 10 10-3 3160 316 100 31 10 Signal duration [µs] Signal duration [µs] (a) PMD for exhaustive and iterative searches.. The total discovery delay of the pure CI-based technique is higher than that of the exhaustive approach. the channel propagation is affected by many factors. in case they contain a stronger path. Figure 2: Trade-off between delay and PMD. considering closer users in Figure 2(b). From the results of Table III. In Figure 3. BS-UE distance d = 35 m. whereas the UE does not need to do any scanning since it knows the BS location and can beamform to it directly). the enhanced CI-based technique.g. with respect to the exhaustive search. even iterative techniques lead to a sufficiently low total discovery delay. We then propose a more sophisticated scheme. vs. according to the analysis we made in Subsection V-A. as it requires an unacceptable total delay with respect to that of the exhaustive scheme. although this is not addressed explicitly in this study. instead. the UE points a beam over the direct path inferred via CI. This requires a total of Ns = 64 slots if three beams are used (the one corresponding to the direct path plus one on each side). B. iterative techniques try to compensate for the lower BF gain in the first phase by collecting enough energy to obtain a sufficiently high SNR. as in NLOS or multi-path scenarios). if we want to guarantee good PMD just for closer users (e. with a consequent delay of approximately Ns · Tsig /φov ' 403 ms. Then. as its gain in terms of needing fewer slots (roughly a factor of 3 in this example) is outweighed by their longer duration (about one order of magnitude). the direct path obtained through CI-based GPS coordinates may be a poorer choice with respect to the exhaustive procedure in which. Note that in these examples the iterative search. As a consequence. the best beam is surely found as it is selected through a complete scan.100 0 Iterative 64×4 Iterative 64×16 Exhaustive 64×4 Exhaustive 64×16 PMD threshold ( = 0. where . when wanting to grant good coverage levels for users at 95 meters from the BS. not correspond to good channel conditions (e. as mentioned earlier. greatly increasing the discovery delay. On the other hand. The reason is that. while. In both algorithms.. in a very dense urban environment. we determine the minimum signal duration to meet the usual requirements of PMD (< 0. 2) For the enhanced CI-based search. is outperformed by the exhaustive approach if a PMD target is imposed.g. an iterative approach is not preferred. where a successive beam refinement is performed: first. (b) PMD for exhaustive and iterative searches. the minimum slot duration to meet the target PMD is Tsig ' 630 µs. signal duration Tsig . despite the lower number of required slots. BS-UE distance d = 95 m. which requires Ns = 32 slots (note that in this case the BS has to span the whole angular space. the UE forms additional beams in some adjacent directions. threshold when the signal duration is around 125 µs. despite requiring longer slots. in smaller cells). we show that exhaustive schemes reach the threshold even when adopting the minimum allowed signal duration Tsig = 10 µs.

[8] C. if compared to exhaustive algorithms. A. thus steering beams through adjacent directions. Available at http://networks. and M. [2] L. However. Nov 2014.” in Proceedings of 21th European Wireless Conference. 3437– 3458. Xiao. S. Rappaport. and J. 12. C ONCLUSIONS AND FUTURE WORKS In this work. Korakis. Sun.” IEEE Communications Magazine.06483. [5] C. S. Krzymien. multipath mm-Wave environment. BS-UE distance d = 95 m.01). Capone. the PMD drops to acceptable levels. White Paper. Rangan. S.org/abs/1511. [7] Q. Panwar. Available: http://arxiv. V. Zhao. vol. pp. Wong. pp. Rangan. no. we have studied. pp. S.g.” in 48th Asilomar Conference on Signals.” May 2014. Rong. pp.” in Proceedings of 50th Annual Conference on Information Sciences and Systems (CISS). 63. Zorzi. Y. and M. LTE. Park. Desai. “Anchor-booster based heterogeneous networks with mmwave capable booster cells. and consequently reduced beamforming gains. analyzed and compared some possible implementations of initial access techniques for the upcoming 5G mm-Wave cellular networks. I. 6. 2016. the refinement is performed. and M. 53. Li. Systems and Computers. vol. and T.01) 10-1 10-2 10-3 3160 316 100 31 10 Signal duration [µs] Figure 3: PMD for exhaustive. “Millimeter wave cellular networks: A MAC layer perspective. [3] M. vol.com/file/28771/5g-white-paper. R. • If a minimum coverage level is to be guaranteed with high probability at relatively large distances (around 100 meters from the BS) and when considering a realistic channel model for a very dense. May 2015. which almost overlaps with the results obtained through the exhaustive search. [14] T. S. when implementing a simple refinement of the direct link. “Context information for fast cell discovery in mm-Wave 5G networks. Chen. in case of ultra-dense cell deployments). “Directional initial access for millimeter wave cellular systems. Qian. . Y. iterative schemes exhibit higher misdetection probabilities in general. N. S.” in IEEE 26th Annual International Symposium on Personal. 12. C. vol. a signal duration of around Tsig ' 150 µs is sufficient to meet the misdetection requirements. vol. Fodor. December 2014. 1926– 1930. June 2014. 136–143. Andrews. Zorzi. Schulz. Sciancalepore.” IEEE Transactions on Communications. Finally. G. Capone. T. Sesia. Hosseini. and R. “Millimeter wave mobile communications for 5G cellular: It will work!” IEEE Access. [4] S. “Millimeter wave channel modeling and cellular capacity evaluation.” IEEE Journal on Selected Areas in Communications. P. Sciancalepore. Giordani. Erkip. Aug 2015. Hosseini. Y. iterative and CI-based searches.” IEEE Wireless Communications. and Mobile Radio Communications (PIMRC). [Online]. vol. “Directional cell discovery in millimeter wave cellular networks. Liu. K. “Key elements to enable millimeter wave communications for 5G wireless systems. Samimi. Korakis. G. Mezzavilla. In this case. 21. Barati. [15] M. Indoor. Rangan. iterative techniques require less time to perform the angular search. we will investigate additional enhanced IA procedures that exploit the Context Information obtained through the µW link. “Obstacle avoidance cell discovery using mm-Waves directive antennas in 5G networks. J. Zorzi. M. Filippini. 2706–2716. M. Whether such reduction in delay. 93–98. [9] C.” CoRR. January 2015. P. S. The main idea is that. Wiley Publishing.” IEEE Transactions on Wireless Communications. in order to overcome the hostile mm-Wave channel propagation conditions. Oct 2015. 14. Wu. 2015. R. 6664–6678. P. because of the wider beams used in the first phase. 335–349. urban. Panwar. Sun. J. Tremolada. S. Samimi. Hu. and A. Wu. Al-Shalash. 2013. Alkhateeb. and M. Rangan. Gutierrez. Barati. S. H. N. “Random access in millimeter-wave beamforming cellular networks: issues and approaches.” IEEE Transactions on Wireless Communications.nokia. no. Sartori. Mezzavilla. S. the simultaneous steering of narrow beams in multiple directions through digital and hybrid beamforming architectures will be considered. Niu. and D. Dec 2013. which is part of our future work. I. 10. B. Soong. M. and F. M.” in IEEE Globecom Workshops (GC Wkshps). “Comparative analysis of initial access techniques in 5G mmWave cellular networks. and E. Caramanis. Jeong. [11] A. [13] A. Shokri-Ghadikolaei. exhaustive procedures are to be preferred because they achieve smaller total discovery delay while meeting the target PMD values (< 0. Liu. and G. has the potential to drastically reduce the PMD. pp. needs to be further investigated by looking in more detail at the cost and accuracy of CI. no. is to be preferred to a sequential-based scan. 1164–1179. we will extend this work to multi-cell multi-user scenarios. T. June 2013. no. vol. Wang. T. A. “Looking ahead to 5G. Filippini. vs. “User association for load balancing in heterogeneous cellular networks. Wei. L. W. As part of our future work. The UMTS Long Term Evolution: From Theory to Practice. 2349–2353. Fischione. 2009. pp. pp. Mayzus. on the other hand. Dec 2015. [6] H.06483 [10] V. 180–185. Baker. 6. Toufik. no. pp. 32. Otherwise. vol. I. and we will study the cost of obtaining CI as well as the performance implications of CI inaccuracy. H. when considering closer users (e. Rappaport. Popovski. at the expense of getting GPS coordinates. directionality should be used also in the initial synchronization-access phase. Ye. R EFERENCES [1] Nokia. • Pure CI-based 64×16 Exhaustive 64×16 Enhanced CI-based 64×16 Iterative 64×16 PMD threshold ( = 0. iterative techniques still present acceptable delay values. pp. Rappaport. and the total discovery delay is almost 50 percent lower than in the exhaustive case. making it possible to drastically reduce the total discovery delay. Barati. Y. abs/1511. Our analysis has indeed demonstrated the following key findings: • There exists a trade-off between IA delay and misdetection probability: on the one hand. 1. Furthermore. and V. 1. signal duration Tsig .. “Initial beamforming for mmWave communications. Yu. S. G. C. Azar. 6. Akdeniz. and H. [12] Q. M. Hu. pp. VI.Misdetection probability 100 The misdetection probability of a pure CI-based approach is higher than for the exhaustive search. [Online]. no.