Cognitive radio adaptation for power consumption minimization using biogeography-based optimization

Abstract

Adaptation is one of the key capabilities of cognitive radio, which focuses on how to adjust the radio parameters to optimize the system performance based on the knowledge of the radio environment and its capability and characteristics. In this paper, we consider the cognitive radio adaptation problem for power consumption minimization. The problem is formulated as a constrained power consumption minimization problem, and the biogeography-based optimization (BBO) is introduced to solve this optimization problem. A novel habitat suitability index (HSI) evaluation mechanism is proposed, in which both the power consumption minimization objective and the quality of services (QoS) constraints are taken into account. The results show that under different QoS requirement settings corresponding to different types of services, the algorithm can minimize power consumption while still maintaining the QoS requirements. Comparison with particle swarm optimization (PSO) and cat swarm optimization (CSO) reveals that BBO works better, especially at the early stage of the search, which means that the BBO is a better choice for real-time applications.

Nature‐inspired optimisation algorithms assisted realisation of green communication via CR:a comparison study

“…The other metric for the performance comparison is optimal generation number at which convergence is achieved. It was observed that in all the 30 independent runs, convergence for WOA was achieved in only four iterations, which is very low as compared with about 200 iterations required by the BBO algorithm in [32]. Although the average optimal generation number for ALO is also 4, it varies between 3 and 5 making ALO relatively less stable in terms of convergence speed.…”

“…i. Shrinking encircling mechanism: This behaviour is achieved by decreasing the value of a as in (32). A is randomly generated in the interval [−a, a] and a is reduced from 2 to 0 over the course of iterations.…”

“…(ii) A novel penalty introduction mechanism using exponential function has been adopted which penalises the particles as per the extent of violation of constraints. (iii) The dimensionality of the parameter adaptation problem in the present work is two times higher than the system reported in [32]. To solve such a higher-dimensional and complex problem, we have investigated the performance of recently proposed and highly efficient NI algorithms, i.e.…”

“…Although Qi et al addressed the power consumption optimisation problem in [32] using a meta-heuristic algorithm, i.e. the discrete version of BBO.…”

“…the discrete version of BBO. However, our proposed system model and the problem solving approach is significantly different; the major contributions and differences of which are as follows: (i) to make the system model more realistic, we have considered the constraint of adjacent channel interference (ACI) caused by secondary user (SU) transmission at PU receiver (PU_Rx) that was missing in the work reported in [32]. (ii) A novel penalty introduction mechanism using exponential function has been adopted which penalises the particles as per the extent of violation of constraints.…”