Performance Evaluation

For analyzing the performance of the proposed DPRA scheme, the DPRA scheme is compared to the optimal method, which obtains the optimal solution by exhaustively solving the mathematical equations given in (3.15)-(3.18). But when performing the consistent allocation for a user, the optimal method will choose the sort of consistent allocation which achieves the maximum system throughput. The DPRA scheme is also compared to one conventional scheme called the efficient and fair scheduling (EFS) algorithm proposed in [39]. The EFS algorithm allocates slots to each service according the order of UGS, rtPS, nrtPS, and BE, where UGS (BE) has highest (lowest) priority. At each slot time interval, it assigns a slot of the selected subchannel to the user with maximum transmission rate on that subchannel. If all the subchannels of a certain slot interval are exhausted, the EFS algorithm will move to the next slot interval and perform allocation slot by slot iteratively. It is an intuitive algorithm and its performance is close to an optimal solution [39], [41].

For simplicity, based on the allocated subchannel, modulation order, and available slots, the DPRA scheme will perform consistent allocation for each service type of the selected user. Sequential slots on the selected subchannel will be allocated for UGS, rtPS,

of users is varied from 4 to 40. Each user is assumed to have voice, video, HTTP, and FTP traffic. The maximum system transmission rate per frame which equals to 7.3728 Mbps can be achieved when the highest modulation order is assigned in each slot. We define the traffic intensity as the ratio of the total average arrival rate of all service types of all users over the maximum system transmission rate. Besides, the average arrival rate for voice, video, HTTP, and FTP is 4.8 Kbps, 64 Kbps, 14.5 Kbps, and 88.9 Kbps, respectively.

Figure 3.2 shows the system throughput versus the traffic intensity. It can be seen that the system throughput of proposed DPRA scheme is close to that of the optimal method; the former is just less than the latter by an amount of 3.58% at traffic intensity 0.8. This conforms to the statement in [41]: the result of the greedy method is close to the optimal solution when the number of user is large. It can also be found that the proposed DPRA scheme achieves higher system throughput than the EFS algorithm, especially at higher traffic load. It is because at high traffic intensity, the proposed DPRA scheme can dynamically adjust the priority values, and the more urgent service will be given a higher priority value and allocated more resource to avoid packet dropping. The DPRA scheme can allocate the radio resource more effectively. On the other hand, the EFS algorithm performs slot by slot allocation using fixed priority scheme and can gain the performance close to optimal solution. The EFS algorithm attains the system throughput as large as that of the DPRA scheme until the traffic intensity exceeds 0.6.

However, due to the lack of dynamic priority, which reflects the urgency of each service, in EFS algorithm, more packets may be dropped and system throughput degrades at high traffic intensity.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0

1 2 3 4 5 6 7

Traffic Intensity

System Throughput(Mbps)

DPRA EFS Optimal

Figure 3.2 System Throughput

Figure 3.3 shows the packet dropping rate of voice traffic. The proposed DPRA scheme, the optimal method, and the EFS algorithm attain the voice packet dropping rates close to zero, which fulfills the voice QoS requirement of 1%. The reason is that these three schemes allocate the voice traffic for UGS to a constant amount of bandwidth and the resource allocation is prior to others.

Figure 3.4 shows the packet dropping rate of video traffic. The video packet dropping rate of the DPRA scheme keeps smaller than the QoS requirement of dropping rate (1%) until the traffic intensity is above 0.9. On the other hand, the video packet dropping rate of the optimal method (the EFS algorithm) keeps guaranteed until the traffic intensity is above 0.7 (0.6). The DPRA scheme designs an appropriate dynamic priority value for each service, which is adjusted frame by frame. According to the QoS requirements of video traffic, the priority value of video packets can be promoted and most video packets can be served in time to avoid discarding. But the optimal method is

granted to the video streaming service is not sufficient enough; the EFS algorithm allocates the system resource to video service right after finishing allocation to voice service.

Figure 3.3 Voice Packet Dropping Rate

Figure 3.4 Video Packet Dropping Rate

Figure 3.5 shows the ratio of unsatisfied HTTP users, which is defined as the number of HTTP users, whose average transmission rate is less than the minimum required transmission rate, over all HTTP users. For the DPRA scheme, a high priority value will be given to the HTTP user if its average transmission rate is going to be lower than the minimum required transmission rate. Therefore the ratio of unsatisfied HTTP users of DPRA scheme keeps close to zero even when the traffic intensity is high, and the minimum required transmission rate can be guaranteed. On the other hand, the optimal method is mainly to maximize the system throughput. Thus sometimes, the minimum required transmission rate cannot be assured when the traffic intensity becomes high. The EFS algorithm is designed with a fixed priority scheme which initially assigns service traffic with priorities according to their QoS requirements. Thus their ratios of unsatisfied HTTP users will become increasing with traffic load due to lack of enough resource allocated for each HTTP user.

Ratio of unsatisfied HTTP users

Figure 3.5 Ratio of Unsatisfied HTTP Users

FTP traffic will be transmitted only when voice, video, and HTTP traffic have already been served. Thus its average throughput is the lowest. For the DPRA scheme and the optimal method, the FTP traffic is also specified with the lowest priority value. But the DPRA scheme (the optimal method) achieves the larger (the largest) system throughput, as illustrated in Fig. 3.2, thus comes the result. The optimal method outperforms the DPRA and the EFS scheme by 67.9% and 75.3% in FTP average transmission rate at the traffic intensity 0.8, respectively.

Figure 3.6 FTP Average Transmission Rate

Figure 3.7 shows the average number of iterations per frame for the proposed DPRA scheme, the optimal method, and the EFS algorithm. Here, an iteration is defined as a search for an optimal pair of user and subchannel from K users and N free subchannels to be allocated with a slot. The DPRS scheme takes a much smaller number of iterations than the EFS algorithm. It is because the DPRA scheme designs a consistent allocation mechanism, where the iteration computation for allocation to a user in each frame takes only one time and the number of slots allocated to the user could be more

than one. Therefore, the number of iterations by the DPRA scheme is greatly reduced, which is much smaller than the total number of slots, N L× , in a frame. On the other hand, the EFS algorithm performs the iteration computation for allocation to a user slot by slot. It searches for an optimal pair of user and subchannel for each symbol. Also, the EFS algorithm may need more than one iterations for each slot allocation to an optimal pair of user and subchannel if there is a power constraint violation. Thus the average number of iterations per frame by the EFS algorithm could be larger than the total number of slots, N L× , in a frame. Moreover, in an iteration, the DPRA and the EFS just search for an pair of user and subchannel and check the power constraint. The complexity of an iteration by the DPRA is almost the same as that by the EFS and equal to (O KN . It can also be found that the number of iterations by the optimal method is ) much lager than that by the DPRA scheme. The optimal method takes a number of iterations more than 23,500 times, while the DPRA scheme needs only 23 iterations, at traffic intensity 0.8.

3.5 Concluding Remarks

In this chapter, a dynamic priority resource allocation (DPRA) scheme which performs consistent allocation is proposed for IEEE 802.16 uplink system with multimedia traffic. The DPRA scheme intends to maximize the system throughput and fulfill QoS requirements. It originally designs a priority value for each service type according to the urgency and QoS requirements of the traffic and dynamically adjusts it frame by frame. Simulation results show that the performance of the DPRA scheme is better than the conventional EFS algorithm, which performs allocation slot by slot.

Besides, benefited from the consistent allocation, the computational complexity of the DPRA scheme is much less than that of the conventional EFS algorithm. Also, the system throughput of proposed DPRA scheme is close to that of the optimal method when traffic intensity is larger, but the computational complexity of the DPRA scheme is much less than that of the optimal solution. Therefore we can conclude that the proposed DPRA scheme can reach throughput maximization and QoS satisfaction with lower computational complexity and transmission overhead. The DPRA scheme would be suitable for real applications.

Chapter 4

A Utility-based TMCR Scheduling Scheme for Downlink MIMO OFDMA Systems

4.1 Introduction

Multiple-input-multiple-output (MIMO) based orthogonal frequency division multiplexing (OFDM) gives a solution for future wireless communication since it can help to achieve high system capacity and provide transmit/receiver diversity for reliable communication link. Resource management for multiuser OFDM (MU-OFDM) systems has recently been investigated [21]-[25], where topics were focused on transmission power allocation, subcarrier allocation, bit allocation, or adaptive modulation and coding (AMC). The design goal is to maximize system capacity, minimize total transmission power, provide fairness, or guarantee QoS requirements.

for the MU-OFDM system to minimize the total transmission power using Lagrangian multiplier [21]. However, the computational complexity is too high to make it feasible.

Yin and Hui studied a so-called low-complexity algorithm, where the total bit rate was maximized, subject to users’ rates and total power constraintsby decoupling an NP-hard combinatorial problem into two steps [22]. The first step determines the required power and the number of subcarriers for each user; the second step then assigns the subcarrier and the number of bit for each user. However, the computation is still too complicated because of the Hungarian algorithm. Bala and Cimini proposed a low-complexity resource allocation algorithm to minimize the total transmit power using Lagrangian multiplier [23]. This algorithm was based on the idea of linearizing the original problem by transmitting data at the same rate on each subchannel. The computational complexity of this simplification method is still too high. Zhang and Letaief presented a two-step reduced-complexity subcarrier-bit-and-power allocation algorithm, which firstly loosens the rate constraints and allocates subcarriers to maximize the throughput, and afterwards adjusts the subcarrier allocation step by step to satisfy the rate constraints [25]. This iterative algorithm still takes too much time to solve the problem, which is infeasible for real-time applications.

Lee, Choi, and Bahk proposed an optimal opportunistic scheduler to maximize the total utility of a wireless system [53]. Neely presented a theory of utility and delay tradeoffs for general data networks with stochastic channel conditions and randomly arriving traffic [54]. Lei et al. proposed a packet scheduling algorithm in the downlink of OFDMA system for mixed services [55]. They mapped the system resource (bandwidth or power) or performance measures (data rate or delay) to the corresponding utility and optimized the established utility system, where the utility function is either concave or convex. They then used the utility theory to solve the scheduling problem. However,

they did not consider QoS requirements of different service types, and they took only the minimum transmission rate or only the delay time as the QoS requirement. The computational complexity is too high as well.

For multiple-input-multiple-output OFDM (MIMO-OFDM) systems, the computational complexity on radio resource scheduling for downlink multiuser increases exponentially with the number of subcarrier, multiuser, transmitting antenna, and receiving antenna. Fuchs, Galdo, and Haardt [26] proposed a low-complexity space–time–frequency scheduling for MIMO systems with SDMA. Hara, Brunel, and Oshima [27] proposed a spatial scheduling with interference cancellation in multiuser MIMO systems. They focused on maximizing the system throughput but did not consider the QoS requirement. Xu, Wang, and Zhang [28] presented a two-step multiuser scheduling algorithm for system throughput maximization with reduced complexity in a downlink MIMO/OFDMA system with transmitting preprocessing. They decoupled the multiuser scheduling problem into frequency and spatial domains. The preprocessing scheme decomposes the multiuser downlink MIMO channel into multiple parallel independent single-user MIMO channels (like OFDMA). However, the number of simultaneously transmitted users is restricted by the number of transmitting antennas.

The computational complexity of the scheduling algorithm is still too high. Hu, Yin, and Yue proposed a low computational complexity scheduling algorithm for a downlink multiuser MIMO-OFDM system in [29]. The more the number of transmitting/receiving antenna and users, the higher the system capacity, which results from space diversity and multiuser diversity. However, the scheduling algorithm did not consider the user demands and QoS requirements; also its symbol by symbol allocation is not suitable for current communication systems.

MIMO-OFDMA system, called cross-layer design of packet scheduling (CDPS) [30]. It serves users by considering fixed priority of service traffic. The priority orders of the service traffic are real time service, non-real time service, and best effect service. The real-time traffic can be served in time at low traffic intensity, while the transmission rate of non real-time traffic is too low to fulfill the requirement rate. Tsai et al. proposed a dynamic priority scheduling scheme for downlink OFDMA/SDMA system, called adaptive radio resource allocation (ARRA) scheme [32]. It gives high priority to urgent users and dynamically adjusts the priority of users frame by frame. However, the ARRA scheme adjusts the priority only depending on the time to expiration but not giving the clear differentiation of real-time service from non-real-time one. This may result in that the real-time service may have high possibility to be served after the non real-time traffic at high traffic intensity.

This chapter proposes a utility-based throughput maximization and complexity reduction (U_TMCR) scheduling scheme for downlink multiuser MIMO-OFDMA systems. The U_TMCR scheme firstly designs a utility function to adjust the priority of a user. The utility function is composed of a QoS monitoring function and a radio resource function, where the QoS monitoring function is to indicate the service priority and the degree of urgency of the user, and the radio resource function is to show the spectrum efficiency if it is used by the user. Subsequently, the U_TMCR scheme formulates the scheduling problem into utility-based optimization equations with an objective to maximize the overall system utility value under two system constraints.

Finally, the U_TMCR scheme proposes a heuristic TMCR algorithm to solve the utility-based scheduling problem. The heuristic TMCR algorithm is a greedy method to maximize the overall system utility value. It contains two functions. The first one is a maximum utility allocation (MUA) function, which finds the most appropriate

subchannel for the user and determines the receiving antenna by a multiple antenna selection (MAS) scheme [56], [57] according to the highest individual utility value. The second one is a consistent allocation (CSA) function, which generalizes the allocation result determined by the MUA function on the following consecutive OFDMA symbols according to the urgency value such that users’ QoS requirements can be fulfilled and computational complexity can be reduced.

Simulation results show that the proposed U_TMCR scheme can achieve the system throughput very close to the optimal solution of the optimization equations by exhaustive search and even higher than those of the ARRA scheme and the CDPS scheme by an amount of about 8% and 21%, respectively. Also, the overall QoS satisfaction degree by the U_TMCR scheme is higher than those by the ARRA scheme and the CDPS scheme. The packet dropping probability of real-time services can be kept lower than the requirement by the U_TMCR scheme, but those by the ARRA and the CDPS schemes would violate the QoS requirement at high traffic intensity. Moreover, the U_TMCR scheme reduces computational complexity. Generally, the total number of allocation trials of the U_TMCR scheme in a frame is reduced by 6.25% ~ 29.2%, compared with the ARRA scheme.

The chapter is organized as follows. The system model of the considered downlink multiuser MIMO-OFDMA system is described in section 4.2. Section 4.3 presents the U_TMCR scheduling scheme. Section 4.4 discusses the performance of the U_TMCR scheduling scheme. Finally, conclusions are given in section 4.5.