Thesis or Dissertation Studies on Power Allocation for Multiple Access with Successive Interference Cancellation

林, 会発  ,  Lin, Huifa

pp.1 - 86 , 2016-03-25 , The University of Electro-Communications
In future wireless communication networks, the number of devices is likely to increase dramatically due to potential development of new applications such as the Internet of Things (IoT). Consequently, radio access network is required to support multiple access of massive users and achieve high spectral efficiency. From the information theoretic perspective, orthogonal multiple access protocols are suboptimal. To achieve the multiple access capacity, non-orthogonal multiple access protocols and multiuser detection (MUD) are required. For the non-orthogonal code-division multiple access (CDMA), several MUD techniques have been proposed to improve the spectrum efficiency. Successive interference cancellation (SIC) is a promising MUD techniques due to its low complexity and good decoding performance. Random access protocols are designed for the system with bursty traffic to reduce the delay, compared to the channelized multiple access. Since the users contend for the channel instead of being assigned by the base station (BS), collisions happen with a certain probability. If the traffic load becomes relatively high, the throughput of these schemes steeply falls down because of collisions. However, it has been well-recognized that more complex procedures can permit decoding of interfering signals, which is referred to as multi-packet reception (MPR). Also, an SIC decoder might decode more packets by successively subtracting the correctly decoded packets from the collision. Cognitive radio (CR) is an emerging technology to solve the problem of spectrum scarcity by dynamically sharing the spectrum. In the CR networks, the secondary users (SUs) are allowed to dynamically share the frequency bands with primary users (PUs) under primary quality-of-service (QoS) protection such as the constraint of interference temperature at the primary base station (PBS). For the uplink multiple access to the secondary base station (SBS), transmit power allocation for the SUs is critical to control the interference temperature at the PBS. Transmit power allocation has been extensively studied in various multiple access scenarios. The power allocation algorithms can be classified into two types, depending on whether the process is controlled by the base station (BS). For the centralized power allocation (CPA) algorithms, the BS allocates the transmit powers to the users through the downlink channels. For the random access protocols, there are also efforts on decentralized power allocation (DPA) that the users select transmit powers according to given distributions of power and probability, instead of being assigned the transmit power at each time slot by the BS. In this dissertation, the DPA algorithms for the random access protocols with SIC are investigated and new methods are proposed. First a decentralized multilevel power allocation algorithm to improve the MAC throughput performance is proposed, for the general SIC receiver that can decode multiple packets from one collision. Then an improved DPA algorithm to maximize the overall system sum rate is proposed, taking into account of both the MAC layer and PHY layer. Finally, a DPA algorithm for the CR secondary random access is proposed, considering the constraint of interference temperature and the practical assumption of imperfect cancellation. An opportunistic transmission protocol for the fading environment to further reduce the interference temperature is also proposed. For the future work, the optimal DPA for the random access with the SIC receiver is still an open problem. Besides, advanced multiple access schemes that aim to approach the multiple access capacity by combining the advantages of the network coded cooperation, the repetition slotted ALOHA, and the SIC receiver are also interesting.

Number of accesses :  

Other information