Energy Efficient Random Transmission Control for Cognitive Radio Systems

Abstract

Cognitive radio technology will allow terminals to access licensed and unlicensed portions of the spectrum. This feature will improve end-user satisfaction and will partially solve bandwidth scarcity problems. However, this opportunistic access implies more transmission attempts and thus higher power consumption. This goes against the energy/power efficient design that underpins modern wireless communication systems. This paper partially addresses this issue by proposing a random transmission policy that is energy-efficient and that provides high throughput gains. To facilitate analysis, a reception model for Rayleigh channels is here proposed that allows the calculation of correct packet reception statistics in the presence/absence of interference between primary/secondary users. The analysis initially focuses on the derivation of the boundaries of two types of trade-off regions: primary vs. secondary throughput, and sum-throughput vs. power consumption. It is observed that secondary transmissions always increase power consumption, and in the case of low interference they always lead to higher sum-throughput at the expense of reduced primary performance. By contrast, in the case of high interference, secondary transmissions can reduce both sum-throughput and primary user performance, thus requiring more complex control. It is shown that the minimum sum-throughput solution is also the boundary of the region where primary/secondary contributions to sum-throughput start to become dominant. An optimum transmission policy is further derived that maximizes sum-throughput while keeping primary/secondary throughput and power consumption under control. Sketches of the trade-off regions show the benefits of the proposed transmission policy.