Electrical and Electronic Engineering - Theses

Permanent URI for this collection

Search Results

Now showing 1 - 1 of 1
  • Item
    Thumbnail Image
    Resource allocation in cognitive radio networks
    LIMMANEE, ATHIPAT ( 2012)
    This thesis focuses on optimal power allocation problems for various types of spectrum-sharing based cognitive radio networks in the presence of delay-sensitive primary links. To guarantee the quality of service in the delay-sensitive primary network, primary user’s outage probability constraint (POC) is imposed such that the transmission outage probability of each primary user is confined under the predefined threshold. We first consider a cognitive radio network consisting of a secondary user (SU) equipped with orthogonal frequency-division multiplexing (OFDM) technology able to access N randomly fading frequency bands for transmitting delay-insensitive as well as delay-sensitive traffic. Each band is licensed to an individual single-antenna and delay-sensitive primary user (PU) whose quality of service is assured by a POC. Assuming full channel state information (CSI) is available at the secondary network, we solve the SU’s ergodic capacity maximization problem subject to SU’s average transmit power, SU’s outage probability constraints (SOC) and all POCs by using a rigorous probabilistic power allocation technique. A suboptimal power control policy is also proposed to reduce the high computational complexity when N is large. Next, we study cognitive broadcast channels with a single-antenna secondary base station (SBS) and M single-antenna secondary receivers (SRs) sharing the same spectrum band with one single-antenna and delay-sensitive PU. The SBS aims to maximize the ergodic sum downlink throughput to all M SRs subject to a POC and a transmit power constraint at the SBS. With full CSI available at the secondary network, the optimal solution reveals that at each timeslot SBS will choose the SR with the highest direct channel power gain and allocate the timeslot to that user. The opportunistic scheduling aspect from the optimality condition allows us to further analyze the downlink throughput scaling behavior in Rayleigh fading channel as M grows large. We then examine a cognitive multiple-access channels with a single-antenna SBS and M single-antenna secondary transmitters sharing the same spectrum band with a single-antenna and delay-sensitive PU. Under an average transmit power constraint in each secondary transmitters and a POC at the primary link, we characterize the ergodic capacity region and two outage capacity regions, i.e. common outage capacity region and individual outage capacity region, in the secondary uplink network by exploiting the polymatroid structure of the problems. Also, the derivation of the associated optimal power allocation schemes are provided. The optimal solutions for the problems demonstrate that successive decoding is optimal and the decoding order can be solved explicitly as a function of joint channel state. Finally, we investigate a transmit power allocation problem for minimizing outage probability of a single-antenna SU subject to a POC at a delay-sensitive and single-antenna PU and an average transmit power constraint at the SU, providing that the SU has quantized channel side information via B-bit feedback from the band manager. By using nearest neighbourhood condition, we can derive the optimal channel partition structure for the vector channel space, making Karush-Kuhn-Tucker condition applicable as a necessary condition for finding a locally optimal solution. We also propose another low-complexity suboptimal algorithm. Numerical results show that the SU’s outage probability performance from the suboptimal algorithm approaches the SU’s outage probability performance in the locally-optimal algorithm as the number of feedback bits, B, increases. Besides, we include the asymptotic analysis on the SU’s outage probability when B is large.