A Novel Wireless Channel Allocation Strategy Along the Railway

Transcription

1 A Novel Wireless Channel Allocation Strategy Along the Railway Shiya Wen, Hao Wu State Key Laboratory of Rail Traffic Control and Safety Beijing Jiaotong University Beijing, P. R. China Abstract Future mobile communications systems demand for higher data rates and better service quality compared to stateof-the-art systems. One way to achieve this ambitious goal is to use relaying which improves spectrum efficiency by adding one or more intermediate nodes support communication pairs. They can be flexibly deployed than additional base stations while the infrastructure costs is being kept lower. Besides, the concurrent usage of different strategies is likely to be realized in the next generation mobile communications systems, such that according on the actual channel conditions the most beneficial strategy can be chosen. In this paper, a novel strategy of wireless channel allocation is proposed. The target system is the celluar mobile communication system along the railway with relay technology. The numerical results show that: it has good performance in the low complexity,such as high stability, high capacity and channel utilization rate. Index Terms relay; channel allocation; utilization rate; capacity. I. INTRODUCTION The relay system, with its many merits, is one of the most promising techniques for the future of mobile communications systems [1] and [2]. The relay-based cellular system largely consists of three elements: a base station (BS), a relay station (RS), and a mobile station (MS). MS can connect to either BS or RS, according to the location of MS. Communication with the MS connected to RS must be performed through the RS. Use of relay in cellular systems were studied in [3] and [4] and [5] and [6]. Most of these works considered solely data traffic. The aim in [3] and [4] was improving the aggregate throughput or spectral efficiency by using relays in broadband systems, while in [5] fair sub-channel allocation algorithms were proposed and in [6] a heuristic was proposed to jointly improve outage probability and throughput. Unlike all these works we distinguish different traffic types and propose a frame-by-frame scheduler, where in each time slot, the time slots, subcarriers and power in a frame are jointly allocated to each transmission. Our algorithm provides proportional fairness for data users, while satisfying delay requirements for real time sessions. Proportional fairness is a good balance between throughput and fairness for data users. We also convert delay requirements of real time users into rate requirements and treat them as constraints in the optimization problem as in our previously proposed resource allocation scheme for a system without RS in [7]. Wireless resource allocation is an important technology in link layer. It plays an indispensable role to increase capacity of wireless networks. Wireless resource allocation put performance of a single user or all users within the district as a criterion in wireless system. It allocated wireless resource to each user. In 4G system, channel allocation and packet scheduling is combination, a consideration of channels quality and business QoS (Quality of Service) requirements. Many theoretical studies have attempted to give the best compromise between frequency utilization efficiency and business performance [8]. The existing common algorithm about wireless channel allocation should balance the throughput and fairness. Literature [9] introduces CoMP-MU-MIMO (Coordinative Multiple Point-Multi User-Multiple Input Multiple Output) schemes and the performance evaluation corresponding uplink system level. Literature [10] introduces a downlink implementation plan of CoMP (Coordinative Multiple Point) systems. With the new wireless business continuing to appear, the QoS also highlights the characteristics of diversity. The LTE (Long Term Evolution) system is required to support different requirements of business QoS. We should consider the throughput and fairness in the design of algorithm of channel allocation. QoS requirements also should be put into consideration in the design for LTE-A(Long Term Evolution-Advanced)system allocation algorithm. The collaborative scheduling algorithm is not perfect, cooperative scheduling does not have an algorithm own advantage and use resource reasonable after relay joining the queue algorithm yet. There still exist a lot of problems to be solved in base station and layer three relay cooperative scheduling, such as interaction of the information between base stations, and how to balance the performance, complexity, amount of feedback, backward compatibility and other aspects. So the research of packet scheduling and channel allocation is a key in the implementation of the LTE-A. The cells along the railway are set in line. The information between mobile station on the train and the network side only is referred to the cell forward and the one backward. But in urban area, we will meet the wireless resource allocation problem when the train is driving in the cell along the railway, and this problem basically involves the current cell and the next cell where the train will arrive. During the whole traveling

2 process the communication link between the train and ground must be guaranteed. The cells along the railway will provide services for users on the train by using the resources, but it will form negative effect to the users in the cell along the railway, such as communication interruption, failure to access and lack of wireless channels. Therefore, a novel strategy of wireless channel allocation is proposed in this paper based on the study of the existing channel allocation scheme, which is a wireless channel allocation scheme based on the cells along the railway, It aims to solve the issue of shortage of resources in the cells along the railway. The numerical results indicate that, the novel strategy could reduce the block probability of voice and improve the channel utilization rate better. II. SYSTEM STRUCTURE A. The existing mechanism of channel allocation (1) FCAS (Fixed Channel Allocation Scheme): The number of current remaining channel is represented as N, and data packet to be transmitted needs M channels. If N > M system will allocate M channels for packets. Otherwise, transmission request of the GPRS packet will be denied. And voice calls will be blocked only in the absence of channel [11]. (2) DCAS (Dynamical Channel Allocation Scheme): We are assuming that establishing a call for data that transmission with a maximum of M channel, according to the current available channel resources. DRA (Dynamical Resource Allocation) will allocate M channels for this call. If the remaining channel number N M, system will allocate M channels; if 0 N < M, system will assign N channels for it; if N=0, this call will be blocked. DRA strategy actually set the minimum number of using channel 1 or 2 as default. Voice calls will be blocked only in the absence of channel. (3) HCAS (Hybrid Channel Allocation Scheme). This strategy makes a combination of the fixed channel assignment and dynamic channel allocation strategy. Each system assigns a fixed set of channels and reserves a group of flexible dynamic channels. Schedule allocating method is mainly based on the situation of known business. These flexible channels will be allocated to the predetermined system to solve the predictable changes of known business. The use of this allocation system, each volume of business is detected on continuous or periodic. Then different business will be allocated to these channels according to these measured values. Therefore, HCAS is the product which combines FCAS and DCAS [12]. Because of the shortage of wireless resource, we can choose the strategy among the following: using FCAS with fixed GPRS package, using the least number of channels that the GPRS package default (DCAS), or existing frequency interference in the actual (HCAS). All these will reduce the efficiency of wireless resource and the utilization ratio of the channel. B. the novel allocation strategy of radio resource In LTE system, sub-channels can be allocated to different users to exploit the channel condition, and hence maximize the achievable data rate. At the meantime, relay aided system also attracts attention in recent years. Relays are often considered as a means to improve the performance of the infrastructure based network by increasing their coverage area and exploiting the spatial diversity. Due to the two hops in the relay aided system, the resource allocation has become an essential component. Therefore, for a relay system along the railway, in order to achieve better performance, the resource allocation is of great interest. We put forward a novel wireless channel allocation scheme based on the mixed cells along the railway. System model is shown as Fig 1, the model includes the cells along the railway, base station in the center of cell and the relay stations at the cell edge. Based on the linear area, the fixed driving direction, and a combination of train index, we propose a novel predictive wireless channel allocation strategy. Specifically, route of driving is fixed, so according to the train index, in a network database we can store a cell list along the railway which is under the control of network side. When a train is ready to start, it needs to be registered in the network, network side can learn the information of the cells where the train will pass according to the train index [13]. Fig. 1: The diagram of cellular model along the railway Wireless channel allocation strategy based on the cell along the railway is shown in Fig 2, there is mobile management equipment on the train. Function of mobile management equipment includes GPS, transceiver and information collecting. Train mobile management equipment sends request to the network side to register train index, the network side can obtain the train index; according to the train index and the data along the cell previously stored, the network side makes cell list along the railway and coding. The mobile management equipment on the train sends the related information of the train to network side periodic, including the speed of train, train s location, distance to next cell, information of train stops as well as wireless resources amount in the prior two cells along the railway. After the data processing, network can get the time T1 when reaching the next area. We preset time threshold value is T2. If T 1 < T 2, network side sends a message to base station which in the next cell, informing it reserved channels resource for the upcoming train passengers. At the same time through the acquired information, we could calculate statistics reserved resource quantity of prior two cells along the railway and send the result back to the next cell, notify the cell to reserve the channel resource reasonably. The BS of the next cell recycles the wireless resource and reserves

3 them for the passengers according to the periodic updating information and the storage data along the railway. In this case, the quality of reserved wireless resources can be confirmed through reservation information feedback of prior two cells along the railway. MS BS1 BS4 RS Requesttoaccess,apply the resource S1 Feedbackinformation that channel is full withthefeedback information, sending access request applywireless resources to BS4 lendwirelesschannel resources4tomsinthe cell along the railway Allocatingwireless channel resource S4 TrainleavethecellwhereisBS1inforsometimes aftercommunicationis over, channel resources released BS4recycle wireless channel resource Fig. 3: Diagram of the channel allocation Fig. 2: Wireless channel allocation process based on the cell along the railway As shown in Fig 1.There are communication needs in the cells where the train is pass. An access request S1 will be sent to the BS1; since the channels of recycle serves passengers on the train, BS1 feedback information indicates that channel resource is full to users, so the users in the cell tries to request the RS5 at the cell edge to access and transmits the feedback information of BS1 to the RS at the same time. RS4 receives access request and feedback information of BS1 and applies for wireless resource in short time, BS4 in the adjacent cell allocating wireless channel resource in time to serve the users in the cell along the railway through the RS4. When train left the cell at some time, BS1 recycles wireless resources for providing normal communication service to the users in the cell along the railway. At the same time, BS4 in adjacent cell recycles resource S4 through RS4. When the train arrives, network side logs off the train index. The interactive process is shown in Fig 3 III. NUMERICAL ANALYSIS AND SIMULATION A. Numerical analysis In interference-limited environment, the SIR (Signal to Interference Ratio) which MS at different location suffers is a very important problem. In this part, we make analysis to the SIR of resource allocation in this system model [14]. Due to borrow channel resource from adjacent cell, interference along the railway is the largest, a link along the railway, we assume interference source number is L. d and γ d are represented as distance of link and coefficient of path loss. P D is the transmitting power on the link by transmitter. d I,1 is the distance that transmitter should consider in the first link from L interference sources,γ I,l is coefficient of path loss that transmitter should consider in the first link from L interference sources, P I,L is transmit power of the first interference source. SIR is given by: Γ(d, γ D, d I,l, γ I,l ) = P D / P L(d, γd ) L [ PI,l / ] (1) P L(d I,l, γ I,l ) l=1 By our analysis, in order to obtain the maximum interference, we assume to produce maximum interference by L interference source at different location. The main interference sources that relay suffered are from adjacent cells, seeing the traditional cellular network structure, the distance from each interference source to the relay is 3R R 0. (R 0 is the distance from relay to mobile node, that is R 0 = 2 3R). When the point to point communication is complete by a two hop link, we consider SIR depends on the smaller one of two link, therefore, SIR can be expressed as the following form: Γ 2 (d F ) = min(γ F M (d F ), Γ BF ) (2) In the formula above,γ F M (d F ) is SIR from MS to RS when the distance from MS to RS is d F.Γ BF is SIR from BS to RS. When we consider the SIR on the link between the relay and the MS, the interference is sent from the position where the distance from RS is 3R R 0, therefore, we can get the SIR function on the relay s link: Γ F M = d F (3R R 0 ) 3.5 (3) We focus the SIR on the link from BS to relay now, this SIR is a constant: Γ BF = ( 2 3 R) 2 6 (3R 2R 0 ) 3.5 (4) Where there is a direct link between the MS and BS, we set the distance from MS to BS is a constant d MB, the SIR in this link is given:

4 3.5 d MB Γ BM = 6 (3R R 0 ) 3.5 (5) System resources includes time slot, frequency and code, the distribution of Erlang B is a kind of typical business model. We can analysis the block rate through Erlang B model. If the resources of system available, we can allocate resource to the new call, but the resource is occupation, the next call will be blocked. There is no queue cache. call reach obey Poisson distribution, business time obey negative exponential distribution. In theory system is in balance state. The block rate of Erlang B is given by: A x / p(x) = X! N i=0 Where, A is the number of calls in average time, N is the number of system total source. Here we could study the spectrum utilization rate. The spectrum utilization rate is defined by the average number of bits per second per hertz in unit area [15]. When we make a calculation of spectrum utilization rate, we use Monte Carlo simulation algorithm, at each simulation, when there is a direct link between MS and BS, we obtained the capacity of the channel: A i i! (6) C 1,E = B CH,E log (1+Γ MB(d BM )) 2 (7) Γ MB (d BM ) is the SIR on the link between BS and MS when the distance from BS to MS is d BM. B CH,E is the channel bandwidth when using the distribution program by relay. Similarly, when MS establishes a two-hop link, the channel capacity can be expressed as: C 2,E (d F ) = B CH,E log (1+min(Γ F M (d F ),Γ BF )) 2 (8) Where d F is the distance from MS to RS. Γ F M (d F )is the SIR on the link between MS and RS, Γ BF is the SIR on the link between BS and RS. We assume that there are N1 direct link between MS and BS in the cells along de railway, and there are N2 two-hop links between MS and RS in the adjacent cells. Therefore, the capacity of cell is: C CELL = N 1 M=1 C 1 (d B ) + N 2 N=1 C 2 (d F ) (9) In this system model, we borrow the channel resource from adjacent cells, so there are three cells in each group, so the spectrum utilization rate of unit area is: η E = 3 C CELL A B T OT (10) Where A is total area each group, B T OT is total bandwidth each group. B. System simulation and analysis At this stage, we will simulate for evaluation the effect of the scheme we proposed. We propose the following parameters: B T OT = 108MHZ, R=500m, bandwidth of each carrier channel is equal, as 200 khz, the transmitting power of the mobile stations is equal to the transmission power of relay. When calculating the spectrum utilization rate we assume that the cell is in full load. 1) simulation of blocking probability: We assume the total bandwidth in each group is 108MHz, so number M of the carrier channel in each group is equal to 540. We assume that there are 27 group channels, so there are twenty sub-channels in each group We make experiment exceeding one thousand times by using Monte Carlo model. In every simulation experiment, users are randomly distributed in the cell. Through computer simulation, we can get the result as is shown in Fig 5. block rate novel strategy with relay traditional scheme in the cell traffic volume Fig. 4: The simulation of blocking probability m=25 m=26 m=27 The horizontal coordinate is the traffic volume and vertical coordinate is blocking rate. We can learn from the Fig 6, when the traffic is small relatively, such as from zero to ten, that is to say the flow density A is between zero and one. The blocking rate of novel allocation strategy is lower than that of traditional channel allocation strategy. In the case of low traffic flow, the novel channel allocation strategy is better than the traditional. But with the increasing of A, such as from one to one point five. The performance of novel allocation strategy is close to the traditional scheme. Even later blocking rate is higher than the traditional mode. Because without channel can be borrowed. On the other hand, there are M channels in each cell. As we can see from the fig the more m is large, the smaller is blocking rate. This is consistent with the general imagination, the more channel numbers, the larger the call numbers of accommodate. But with the increasing of A, such as from one to one point five, the density of call reach is greater than the density of call release. Due to the observation for a long time, even if the m point increases a little, It s no use to such a long time (it will be done soon,then continue to block ). But in the case

5 of lower density, it is the most effective way to reduce the blocking rate. Capacity bits/s/hz traditonal scheme without relay scheme with relay novel strategy SNR in db Fig. 5: Three distribution draw of spectrum utilization 2) Simulation of the spectrum utilization rate per unit area: The horizontal coordinate is the value of spectrum utilization rate. And vertical coordinate is SNR (signal to noise ratio). We calculate the average spectrum utilization rate. Results show that: average spectrum utilization rate of novel scheme based on the cell along the railway is 5.30 bits/sec/hz/km 2 ; average spectrum utilization rate of the scheme in the cell based on relay is 4.02 bits/sec/hz/km 2 as well as in the cell along the railway. Average spectrum utilization rate of traditional scheme is 2.38 bits/sec/hz/km 2. In all, the performance of novel strategy is optimal. Spectrum utilization rate improves 31 percent than the scheme with relay. IV. CONCLUSION In a view of performance, Relay channels improve the performance through spatial diversity by using additional paths between source and destination [16]. the proposed novel wireless channel allocation strategy according to the train index and mobile management equipment on the train to allocate the wireless channel predicatively. The numerical results show that: in the low complexity, it has good performance that was proven by computer simulations. The novel scheme has a high stability as well as a high capacity and channel utilization rate. It also improves the performance of SIR in cells. REFERENCES [1] N. S. Nourizadeh. H and T. R. Daly, Performance evaluation of cellular networks with mobile and fixed relay station. IEEE Vehicular Technology Conference.64th., [2] C. H.-F. M. Mehul., The capacity of several new classes of semideterministic relay channels[j], IEEE TRANSACTIONS ON INFORMA- TION THEORY., pp , 2011,10. [3] R. P. T. Imic, D. C. Schultz and P. Wienert., Capacity of a relaying infrastructure for broadband radio coverage of urban areas. IEEE 58th IEEE Vehicular Technology Conference., 2003-Fall. [4] D. C. S. S. J. A. K. Park, C. G. Kang and J. Ihm., Relay-enhanced cellular performance of ofdmatdd system for mobile wireless broadband services. ICCCN 2007, In Proc. of 16th, pp , 2007,8. [5] O. Oyman., Ofdm2a: A centralized resource allocation policy for cellular multi-hop networks. 40th Asilomar Conference on Signals, Systems and Computers,, pp [6] M. Kaneko and P. Popovski., Adaptive resource allocation in cellular ofdma system with multiple relay stations. Proc. of IEEE 65th VTC, pp , 2007-Spring. [7] J. A. T. Girici, C. Zhu and A. Ephremides., Proportional fair scheduling algorithm in ofdma-based wireless systems with qos constraints. 12th International OFDM Workshop (Inowo 07),Hamburg, German, [8] S. A. D. R. G. M., Wavelet-based downlink scheduling and resource allocation for long-term evolution cellular systems. IET COMMUNI- CATIONS., pp , [9] J. L. G. L. C. C. Dajie. Jiang, Qixing. Wang, Uplink coordinated multipoint reception for lte-advanced systems[j], Wireless Communications,, [10] X. Y. L. C. Jianchi. Zhu, Xiaoming. She, A practical design of downlink coordinated multi-point transmission for lte-advanced[j], Vehicular Technology Conference, 2010,71. [11] B. G. Logrippo, Understanding gprsthe gsm packet radio service. Computer Networks, vol. 34, pp [12] M. S. B. M. ErmelT. MtillerJ. Schiller, Performance of gsm networks with general packet radio services, Performance Evaluation, vol. 48, pp , May [13] M. K. Kim and H. S. Lee., Radio resource management for a two-hop ofdma relay system in downlink. Computers and Communications,ISCC 2007., pp , [14] Z. A. H. C. L. Jiang, On the achievable diversity multiplexing tradeoff for the optimal time allocation in the two-way channel, IEICE TRANS- ACTIONS ON COMMUNICATIONS., pp , [15] H. Hu and H. Yanikomeroglu, Range extension without capacity penalty in cellular networks with digital fixed relays, IEEE globecom 2004, pp [16] V. V. Y. Liang and H. Poor, Resource allocaiton for wireless fading relay channels: Max-min solution, IEEE Trans. Inf. Theory,, vol. 53, pp ACKNOWLEDGMENT This paper is supported by the State Key Laboratory of Rail Traffic Control and Safety(Contract No. RCS2010ZZ004), Beijing Jiaotong University. This paper are also supported by Program for Changjiang Scholars and Innovative Research Team in University under Grant No. IRT0949 and the Joint State Key Program of the National Natural Science Foundation of China and the National Railway Ministry of China (Grant No ). The authors would like to thank the anonymous reviewers for their helpful comments that improved this paper.