Distributed Scheduling in Wireless Multiple Decode-and-forward Relay Networks
- PDF / 1,020,315 Bytes
- 14 Pages / 595.224 x 790.955 pts Page_size
- 3 Downloads / 227 Views
Distributed Scheduling in Wireless Multiple Decode-and-forward Relay Networks Zhou Zhang1 · Ye Yan1 · Tongtong Wang1 · Zuohong Xu2
© Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract In this paper, we study the distributed DOS problem in wireless multiple relays networks. By formulating cooperative scheduling as an extended multi-stage stopping problem, an optimal strategy is proposed achieving the maximal average network throughput for distributed channel access. The strategy is in a tri-level threshold structure, and the optimality is rigorously proved. Close-form expression of the maximal average throughput is also derived, and easy implementation is presented with low complexity facilitating network operation. Furthermore, numerical results are provided verifying the strategy’s effectiveness. Keywords Distributed opportunistic scheduling · Joint time and relay diversity · Optimal stopping theory
1 Introduction With explosively increasing demands on enhanced quality of service, conventional communication network faces serious challenges. In it, the medium access control layer and the physical layer are independently designed. By nature that the wireless medium is shared by multiple users and severe channel fading is experienced, a cross-layer design concept, channel-aware scheduling is thus motivated. By letting the MAC layer be aware of physical layer information, users can make scheduling decision depending on channel quality, and be scheduled if links are good. Recently, existing researches have drawn much attentions on centralized scheduling [1, 2]. In these works, a controller collects channel state information (CSI) of all users and schedules those of best channels to access. By doing this, network throughput increases significantly. Nevertheless, research on distributed opportunistic scheduling (DOS) is still in its initial stage. Main hardness lies that users do not know CSI of all users, and thus can not decide when to access the channel in an efficient manner. As a milestone to resolve such difficulties, ordered scheduling in ad hoc Zhou Zhang
[email protected] 1
Tianjin Artificial Intelligence Innovation Center (TAIIC), Tianjin, China
2
National University of Defense Technology, Hunan, China
network of single hop was addressed in [3] based on optimal stopping approach, where implementation is easy due to pure threshold strategy. Extended from it, the problem of opportunistic scheduling under interference channels allowing multiple user simultaneous access is investigated in [4], and the problem with time constraints for real-time service is studied in [5]. In a distributed network, cooperative communication is in great potentials for enhanced scheduling [6–10]. By utilizing cooperation diversity, network performance such as transmission reliability and capacity is further improved. In recognition of these benefits, increasing interests have been witnessed towards cross-layer design for MAC and physical layer. To say a few, DOS in a wireless amplifyand-forward relay
Data Loading...
