Modeling and Characteristics of Wireless Link for HAPS TD-CDMA System

Transcription

1 Journal of Residuals Science & Technology (JRST) Volume 14 No. 3, 2017 doi: /jrst Modeling and Characteristics of Wireless Link for HAPS TD-CDMA System Guan Mingxiang *1, Zhuang Peican 1, Wang Le 1 1 School of Electronics Communication Technology, Shenzhen Institute of Information Technology, Shenzhen, , China *1 gmx2020@126.com Abstract An information system formed by HAP (High Altitude Platform) will be a new generation-system for the wireless communications and HAPS (HAP Station) communication system combines the advantages of both terrestrial and satellite communication systems and avoids, to different extents, their disadvantages. The scope of this paper is to evaluate the link performance of HAPS CDMA air interface. The simulation approach has, with respect to others methods (e.g. hardware prototypes, analytical evaluations), some advantages like lower costs, accuracy, inherent flexibility, etc. The link performance is the necessary input for system capacity and coverage evaluations which are definitely the most important aspects for a telecommunication operator. Keywords HAPS; TD-CDMA; Wireless Link; Performance Introduction Mobile communication technologies have developed rapidly in the past two decades. The fundamental problem of mobile communication is finding ways to enable people to communicate by all means possible, whenever and wherever. The increasing demand for broadband mobile communications has led to the successful and rapid deployment of both terrestrial and satellite wireless networks. Besides the high data rates, current wireless networks can be inexpensive, support reconfigurability and provide time- and space- varying coverage at low cost. Recently, a novel form of mobile communication called the high-altitude platform station (HAPS) has emerged [1 4]. In general, a HAPS network is composed of quasi-static high-altitude platform stations at low altitude ( Km), with a certain payload, and with a residence time of 5 10 years. In the near-earth space, HAPS adopts a stable communication platform as a microwave relay station and forms into a communication system with ground control units, access equipment, and various wireless users. HAPS implement the integrated satellite-ground networking and support a separate ground networking, as shown in Fig. 1. As can be seen, the communication platform is synchronized with the rotation of the earth, and it can reside in space for a long time. HAPS utilize sound radio transmission features and implement the communication connections between users, platforms, and satellites. It also possesses the following advantages: layout flexibility, wide-ranging applications, low cost, security, reliability, and so on [5-8] The HAPS-based information system is a new generation wireless communication system, which combines terrestrial and satellite communication systems under the principle of fostering their strengths and circumventing their weaknesses. This system is widely recognized in the communication fields. In addition, within its coverage, the HAPS is still allowed to use cellular network communication [9-10]. At the moment 3GPP (Third Generation Partnership Project) is focusing on the definition of the terrestrial segment of the UMTS system (the so called UMTS Terrestrial Radio Access, UTRA), based on two different solutions: W- CDMA and TD-CDMA air interfaces [11]. The former has been selected for the paired frequency band allocated to UMTS, while the second will be deployed in the unpaired frequency band. Both W-CDMA and TD-CDMA solutions are under continuous development in 3GPP. In this paper, a link level simulator for HAPS TD-CDMA is described. The link level simulator can be used to evaluate the radio performance of HAPS TD-CDMA system in 281

2 Volume 14 No. 3, 2017 Journal of Residuals Science & Technology (JRST) different scenarios. Actually, when this simulator is being developed, it is used to analyze the communication performance of high speed train [12].Taking into account that the link level simulator allows only the simulation of the physical layer, it is necessary to consider some emulation of system level simulations in further studies. a) SATELLITE-GROUND-HAP INTEGRATED NETWORKING b) GROUND-HAP INTEGRATED NETWORKING FIGURE 1. HAPS COMMUNICATION NETWORK ARCHITECTURE Analysis of Wireless Link Modeling for HAPS CDMA Communication Propagation models aim at predicting the average received signal power at a given distance from the transmitter (large-scale propagation models), as well as the fluctuations of the received power over very short travel distances (a few wavelengths) or short time durations in the order of seconds (small-scale propagation or fading models). In built-up areas, fading occurs because there is no line-of-sight path between the transmitter and the receiver. However, even when a line-of-sight exists, multipath, which creates small-scale fading effects, occurs due to reflections from the ground or surrounding structures. It is worth mentioning that even for the case of a fixed receiver, the received power may fade due to the movement of the surrounding objects. In transmission between a HAP and a ground terminal, propagation often takes place via many paths. A significant fraction of the total energy arrives at the receiver by way of a direct wave. The remaining power is received by way of a specular ground reflected wave and the many randomly scattered rays which form a diffuse wave. Therefore, a signal is received from a number of different paths. The signals of the different paths are all replicas of the same transmitted signal but with different amplitudes, phases, delays and arrival angles. Adding these signals at the receiver may be constructive or destructive. Compared to wireless terrestrial links, HAP links have more favourable propagation characteristics. In wireless terrestrial systems, the received power decays as a function of the transmitter-receiver distance raised to a power of 4. Additionally, the Rayleigh distribution is commonly used to describe the small-scale fading envelope. In HAP links, there exists a dominant signal component such as a line-of-sight path, and the small-scale fading envelope 282

3 Journal of Residuals Science & Technology (JRST) Volume 14 No. 3, 2017 distribution is Ricean. The main parameters of terrestrial wireless and HAP links are illustrated in Table 1. A critical parameter is the Ricean factor K, which is defined as the ratio of the dominant component to the scatter contribution. Typically, the range of K is 0-20 db, and the larger its value, the higher the energy gain in HAP-based systems compared to terrestrial ones, where K is close to zero (Rayleigh fading). TABLE 1 BASIC WIRELESS TERRESTRIAL AND HAP CHANNEL PARAMETERS Parameters Terrestrial HAPS Fast Fades Distribution Rayleigh Ricean Dynamic Range in a Cellbased System db(40-50 db due to propagation-induced difference and db due to fading) db(2db due to propagation-induced difference and db due to fading) HAPS CDMA system is also deigned to mitigate the multi-path effects in wideband environment by using equalization and advanced multi-user receiver. However, as the delay spread increases, the performance of the HAPS CDMA system degrades due to the corruption of the orthogonal relationship among codes. In addition, the complexity of the receiver may increase under this condition. On the other hand, in HAPS CDMA system resistance to the multi-path interference results from the increased symbol duration for each carrier and from use of a cyclic prefix, called guard period, preceding each symbol. In general, the HAPS CDMA system provided performance gain comparing to the territorial CDMA system with similar receiver complexity in a strong multipath environment. Regarding channel modeling in HAP systems, The small-scale fading for a communication link at 2GHz between a terrestrial user (fixed or mobile) and a HAP was carried out in the presence of the paper. Due to the fast fades distribution of Case3, Vehicular A and Rician channel respectively, the simulation was based on a model for terrestrial links which was extended to the HAP-based system. System Model 3.1 Transmitter The transmitter performs the following functions: Cyclic Redundancy Code (CRC) and Tail Inserting; coding (convolutional or Turbo); Rate matching and interleaving; mapping; Modulation and spreading; Burst Formatting: midamble and guard interval insertion. The CRC provides the indication in the end of each data block to see whether there are error(s) within the block. To meet the length requirement in the following processing, the Tail Inserting is necessary. The channel coding includes convolutional coding (rate or 1/3 and constraint length 9) for voice service and Turbo Coding for high quality data services. The following operations are frame equalizations, 1st interleaver and frame segment. The rate matching and interleaving punctures and reorders the data sequences. The channel mapping maps the data stream into the corresponding physical channels or transport channels. The modulation scheme is a simple QPSK scheme for general data rate services. QAM16 is also applied and correspondingly the bit rearrangement is considered. The spreading adopts a unique spreading factor of 16 for downlink communications or HSDPA and a variable spreading factor ranging from 1 (no spreading) to 16 for uplink communications. The spreading weighting and scrambling are also involved. There are three types of burst which can be selected depending on the environment. For Burst type 1 and Burst type 2, the guard interval length is equal to 96 chips and for Burst type 3 the guard interval length is 192 chips. A midamble is inserted in order to allow the channel impulse response estimation at the receiver. The three burst structures are depicted in Fig. 2 through 4 respectively in case of a spreading factor Q equal to 16. In our given simulations, only Burst type 1 and Burst type 2 scenarios are considered but the simulator also supports the Burst 283

4 Volume 14 No. 3, 2017 Journal of Residuals Science & Technology (JRST) type 3 scenarios. 976 chips Midamble 512 chips 976 chips GP 96 chips 666.6us FIGURE 2. BURST 1 STRUCTURE (SPREADING FACTOR Q=16), GP=GUARD PERIOD 1104 chips Midamble 256 chips 1104 chips GP 96 chips 666.6us FIGURE 3. BURST 2 STRUCTURE (SPREADING FACTOR Q=16), GP=GUARD PERIOD 976 chips Midamble 512 chips 880 chips GP 192 chips 666.6us FIGURE 4. BURST 3 STRUCTURE (SPREADING FACTOR Q=16), GP=GUARD PERIOD The type 1 burst adopts a longer midamble (512 chips against 256). For this reason burst 1 is more suitable for environments characterised by higher delay spread and for the uplink where even QMAX (QMAX=maximum spreading factor=16) different impulse responses have to be estimated in parallel. The type 2 burst can be adopted in downlink because only one impulse response have to be estimated by the mobile station. Different midamble sequences are assigned to adjacent cells in order to implement right channel estimation. In the same cell, a different midamble shift of kw (k=0,1,,k-1) chips is assigned to each user in order to allow the base station (BS) or mobile station (MS) to estimate the different impulse responses as shown in Fig. 5. P elements P-1 elements Use K Use K-1 Use K-2 W 2W Base code with period P Use 1 (K-1)W L m FIGURE 5. MIDAMBLE GENERATION 3.2 Model The propagation channel is made up of a Rayleigh multipath fading generator, followed by an AWGN generator. As for the TD-CDMA access scheme, the propagation environments defined by 3GPP: Case1, Case 2 and Case 3 are available as shown in Table 2. In both downlink and uplink, an FFT algorithm is used to estimate the channel coefficients. 284

5 Journal of Residuals Science & Technology (JRST) Volume 14 No. 3, 2017 TABLE 2 PROPAGATION CONDITIONS FOR MULTI PATH FADING ENVIRONMENTS 3.3 Receiver Delay /ns Case 1 speed 3 km/h Mean Power /db Delay /ns Case 2 speed 36 km/h Mean Power /db Case 3 speed 120 km/h Delay /ns Mean Power /db The receiver performs the Joint Detection of the user signals in each slot and the processing functions (e.g. deinterleaving, channel decoding, etc.) necessary to decode the user data sequences. The Joint Detection (JD) techniques are used in CDMA systems in order to perform simultaneously equalisation and signal separation. In Fig. 6 and Fig. 7 the block diagram of the simulated radio chains are shown for downlink and uplink respectively. Source CRC & tail insertion coding Puncturing or unequal repetition Interleave Midamble generator Map Data Modulation & spreading Filter TX Power control command Fading channel estimation AWGN BLER counter CRC & tail removal decoding Depuncturing or repetition removal Deinterleave Demap Joint detection Filter RX FIGURE 6. BLOCK DIAGRAM OF THE HAPS TD-CDMA DOWNLINK RADIO CHAIN Source CRC & tail insertion coding Puncturing or unequal repetition Interleave Midamble generator Map Data Modulation & spreading Filter TX Other users Power control command Fading channel Fading channel estimation AWGN AWGN BLER counter CRC & tail removal decoding Depuncturing or repetition removal Deinterleave Demap Joint detection Filter RX Filter RX FIGURE 7. BLOCK DIAGRAM OF THE HAPS TD-CDMA UPLINK RADIO CHAIN For HAPS TD-CDMA scenarios, the block diagram is similar to the downlink scenarios except that the block is compose of several time slots and the processing block size correspondingly becomes larger. The ARQ (Automatic Repeat request) scheme is adopted. This simulator can support Equal Gain Combining and Chase Combining to improve the transmission quality. Different Modulation and Coding Schemes (MCS) are deployed to adapt to 285

6 Volume 14 No. 3, 2017 Journal of Residuals Science & Technology (JRST) different services. The adopted MCS in the simulator is shown in Table 3. Because the TD-CDMA system simulation need too long time to get the results in all scenarios, we only give the AWGN results for these four cases. However, the extension to other frequency-selective channels is straightforward. TABLE 3 MODULATION AND CODING SCHEME OF HSDPA Parameters MCS1 MCS2 MCS3 MCS4 Modulation QPSK QPSK 16QAM 16QAM Information bit payload Coding rate R=3/4 R=3/4 R=3/4 R=9/16 Code number Timeslot number Spread factor Date rate/kbps Link Difference between Territorial CDMA System and HAPS CDMA System The coverage performance of both territorial CDMA system and HAPS CDMA system are evaluated. The link budget of both uplink and downlink are analyzed and the maximum allowable path loss is used as the performance metric. 4.1 Assumptions In the downlink link budget calculation, a single cell with no inter-cell interference is assumed and all the transmit power is allocated to a user with maximum theoretical downlink throughput. The results indicate the maximum allowable path loss for the highest data rate service in single cell environment. On the other hand, uplink coverage is normally limited by the mobile transmission power in cellular system. The maximum range for a cell to provide service to a low data rate user is calculated in the uplink link budget. The common assumption and parameters are shown in table 4. TABLE 4. LINK BUDGET ASSUMPTIONS Assumptions Base Station User Terminal Cable, connector and duplexer losses(db) 2 0 Body losses(db) 0 0 Receiver noise figure(db) Antenna gain(db) 18 2 Receiver Noise density(dbm/hz) Link Budgets for Territorial CDMA System Receiver Noise Power(dBm) In the uplink analysis, a low data rate service of 31.4kbps is assigned to the user, in which the spreading factor is 16, two codes and one time slot is assigned to the user. The required Eb/N0 is 4.2dB. Table 5 shows an example link budget for territorial CDMA system. It shows that the allowed propagation loss for downlink and uplink are and 140.1dB respectively. This indicates that if high data rate service is provided in the system, the cell range is limited by downlink rather than uplink. This observation is opposite to the normal situation of low data rate voice traffic cellular environment. 4.3 Link Budgets for HAPS CDMA System The link budget analysis is derived based on the assumptions and the bearer specification. in HAPS CDMA system, a particular time period, the whole spectrum is allocated for a single user in downlink and multiple users can be transmit to the base station simultaneously in uplink with part of the whole allocated spectrum. The total number of sub-channels is 32. If a user is only assigned to single sub-channel, then 1/32 of the allocated bandwidth is reserved to uplink user. Since only a small section of bandwidth is assigned to that user in each time slot, the full 286

7 Journal of Residuals Science & Technology (JRST) Volume 14 No. 3, 2017 transmit power can be used in one sub-channel and 15dB sub-channelization gain is added to the uplink budget. In addition, 5MHz spectrum is allocated for both uplink and downlink and the corresponding spectrum usage ratio is 1:4. In order to mitigate the multi-path effect, guard period is inserted in the frame to maintain the orthogonal among sub-carriers. Table 6 Similar to HAPS CDMA system system, the cell range is limited by the downlink with high downlink data rate. The allowable propagation loss for the cell is and 138.2dB for downlink and uplink respectively. TABLE 5. LINK BUDGET FOR TERRITORIAL CDMA SYSTEM Parameters Downlink Link Budget Uplink Link Budget Transmitter Maximum Tx power per code(dbm) Antenna gain(dbi) Tx cable loss/body loss(db) EIRP(dBm) Receiver Thermal noise density(dbm/hz) Receiver noise figure(db) Receiver noise density(dbm/hz) Receiver noise power(dbm) Interference margin(db) Total noise + margin(dbm) Processing gain(db) Require Eb/No(dB) Receiver sensitivity(dbm) Antenna Gain(dBi) Rx cable loss/body loss Maximum path loss(db) Log-normal fade margin(db) Allowed propagation loss for cell range(db) Simulation Results and Performance Analysis Path-loss prediction is the under laying requirement of any system analysis, especially of any HAPS CDMA system simulation and coverage prediction. Most simple Path-loss models use HATA formulas. Such propagation models are statistical and do not take into account the ground precisely. Shadowing effects are generated by a log-normal law with zero mean and a given standard deviation (typically 6dB in rural areas). Given the de correlation square decomposition, path-loss and shadowing samples are only computed at the centre of de correlation squares. If samples between the two positions are needed, the value will be evaluated by interpolation between consecutive values. UMTS system performance is extremely sensitive to path-loss. A few db error in intercept prediction can result in large variation in CDMA capacity and coverage. As a consequence, more complex propagation models with terrain parameters are beneficial, based on data maps. In addition, tuned propagation are sometimes included into simulators. The effect of long term fading can be modeled by adding a random value to the path-loss calculated for the mobile position. The random value is extracted from the probability distribution assumed as the shadowing distribution (e.g. log-normal distribution). Links between the mobile and base stations located in the same site are assigned the same shadowing value. Usually, uplink and the downlink are supposed to be perfectly correlated, thus the same shadowing figure is adopted for both. The shadowing is assigned to the mobile randomly at call set-up or after mobile placement in monte carlo snapshot and eventually updated periodically during the call in time based simulations. In this section, we present the simulation analysis for the high data rate downlink service under various channel 287

8 Volume 14 No. 3, 2017 Journal of Residuals Science & Technology (JRST) conditions. Figure8, Figure9 and Figure10 show the performance impact of various mobility conditions in Case3, Vehicular A and Rician channel respectively. The performance indication is defined as the required Ior/Ioc for a downlink channel to achieve BLER of 0.1. The Case 3 and Vehicular A channel are the reference propagation channel models defined in the 3GPP specification. In high vehicular circumstances, the high trains often run in rural areas, where LOS component appears along the track side about several kilometers from the base stations. The Rician model is also considered in the study for this scenario. The Rician distribution is often described in terms of the K factor, which is defined as the ratio between line of sight (LOS) power and scattered waves power. The K factor of 10dB is used in all the Rician channel simulation, which indicates significant LOS component available and this is usually for conventional base stations built along the Parameters TABLE 6. LINK BUDGET FOR HAPS CDMA SYSTEM Transmitter Downlink Link Budget Uplink Link Budget Tx power per subcarrier(dbm) 8-8 Antenna gain(dbi) Tx cable loss/body loss(db) EIRP(dBm) 24-6 Receiver Thermal noise density(dbm/hz) Receiver noise figure(db) Receiver noise density(dbm/hz) Receiver noise power(dbm) Interference margin(db) Sub-channelinization gain(db) Require SNR(dB) Receiver sensitivity(dbm) Antenna Gain(dBi) Tg/Tb Rx cable loss/body loss Maximum path loss(db) Log-normal fade margin(db) Allowed propagation loss for cell range(db) FIGURE 8. PERFORMANCE COMPARISON OF HIGH DATA RATE DOWNLINK SERVICE IN CASE 3 CHANNEL 288

9 Journal of Residuals Science & Technology (JRST) Volume 14 No. 3, 2017 FIGURE 9. PERFORMANCE COMPARISON OF HIGH DATA RATE DOWNLINK SERVICE IN VEHICULAR A CHANNEL FIGURE 10. PERFORMANCE COMPARISON OF HIGH DATA RATE DOWNLINK SERVICE IN RICIAN CHANNEL The performances of 144kbps are consistently worse than those of 384 kbps. This is mainly due to the mapping between CCTrCH and physical channels. It is configured to have 9 code channels in one time slot compared to 5 and 8 codes in 64 kbps and 384 kbps. In addition, its turbo code interleaver size which is directly impact the BER/BLER performance is smaller than that of 384 kbps. Conclusions The wireless link simulation analysis for HAPS CDMA system is presented in this paper. In this paper, we construct a link level HAPS TD-CDMA simulator for downlink, uplink communications. The simulation results meet the requirement of 3GPP standard, which confirm the validity of this simulator. ACKNOWLEDGMENT This paper is supported by the national natural science foundation of China ( ), the Guangdong Province higher vocational colleges & schools Pearl River scholar funded scheme (2016), project of Shenzhen science and technology innovation committee (JCYJ ). The author would like to thank the editor and the anonymous reviewers for their contributions that enriched the final paper. REFERENCES [1] Yufeng Wang,M. C. Erturk, Jinxing Liu.Throughput and delay of single-hop and two-hop aeronautical communication networks. Journal of Communications and Networks,2015,17(1): [2] A. Ibrahim, A. S. Alfa.Using Lagrangian Relaxation for Radio Resource Allocation in High Altitude Platforms. IEEE Transactions on Wireless Communications, 2015,14(10):

10 Volume 14 No. 3, 2017 Journal of Residuals Science & Technology (JRST) [3] Feihong Dong, Yuanzhi He,Xionglin Zhou.Optimization and design of HAPs broadband communication networks. The 5th International Conference on Information Science and Technology, 2015: [4] Xuyu Wang. Deployment of high altitude platforms in heterogeneous wireless sensor network via MRF-MAP and potential games. IEEE Wireless Communications and Networking Conference (WCNC): NETWORKS, 2013: [5] Mohammed, A., Mehmood, A., Pavlidou, F., Mohorcic, M. The role of high-altitude platforms (HAPs) in the global wireless connectivity. Proceedings of the IEEE, 2011, 99(11): [6] Yiming Liu, Grace, D, Mitchell, P.D. Exploiting platform diversity for GoS improvement for users with different High Altitude Platform availability. IEEE Transactions on Wireless Communications, 2009, 8(1): [7] Feihong Dong, Yuanzhi He, Haitao Nan.System Capacity Analysis on Constellation of Interconnected HAP Networks. IEEE Fifth International Conference on Big Data and Cloud Computing, 2015: [8] Shufeng Li, Grace D, Yanchen Liu, Jibo Wei, Dongtang Ma. Overlap area assisted call admission control scheme for communications system. IEEE Transactions on Aerospace and Electronic Systems, 2011, 47(4): [9] Guan Mingxiang, Yuan Fang, Guo Qing. Performance of Coverage and Wireless Link for HAPS Communication. IEEE International Conference on Wireless Communications and Signal Processing, 2009: 1-4 [10] E. B. Lima, S. A. Matos, J. R. Costa. Circular Polarization Wide-Angle Beam Steering at Ka-Band by In-Plane Translation of a Plate Lens Antenna. IEEE Transactions on Antennas and Propagation, 2015,63(12): [11] 3GPP Radio specifications (25.2 series) [12] Uykan Z., Koivo H. N. Porportional power control algorithm for time varying link gain in cellular radio systems. IEEE Transactions on Communications, 2006, 55(2): Guan Mingxiang is currently a professor at the School of Electronic Communication Technology, Shenzhen Institute of Information Technology. He received his bachelor of Electronic engineering degree from Harbin University of Science and Technology, China, in 2002 and his master and PhD of Communication Engineering degrees from Harbin Institute of Technology, China, in 2004, 2008 respectively. His main interests are in the areas of wireless communication, HAPS communication and network communication research. Zhuang Peican is currently a Engineer at the Prospective National Key Vocational College Construction Office, Shenzhen Institute of Information Technology. He received his bachelor and master of Mechatronic Engineering degrees from Shantou University, China, in 2002, 2005 respectively. From 2005 to 2010, he is a senior engineer engaged in the wireless network performance optimization in Huawei technical service co., LTD, His main interests are in the areas of wireless communication, network communication research. Wang Le is currently a lecture at the Department of Electronic Communication Technology, Shenzhen Institute of Information Technology. She received her bachelor, master and PhD of Communication Engineering degrees from Harbin Institute of Technology, China, in 2001, 2004, 2010 respectively. Her main interests are in the areas of wireless communication and satellite communication research. 290