Date of Award

3-26-2015

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Information Science and Systems Engineering

First Advisor

Seshadri Mohan

Second Advisor

Lifeng Lai

Abstract

Spurred on by the ever increasing demand for bandwidth and the shortage of radio spectrum, researchers are faced with numerous challenges to advance the new technologies. At the same time, spectrum is under-utilized and not fully exploited most of the time in most places. Cognitive radio (CR) is an emerging technology that offers a novel approach to solve the problem of spectrum under-utilization. Specifically, CR offers an approach to share the spectrum through dynamic spectrum access (DSA). In particular, the secondary users (SUs) are allowed to use the communication system without interrupting the primary users (PUs). Spectrum sensing is the tool to identify the unoccupied frequency bands. While minimum detection errors are required, detection delay is also crucial for developing optimal sensing strategies. Detection errors and sensing time are inversely related and hence a tradeoff should be struck between them. The overall goal is to maximize the throughput of CR users. For the problem of spectrum sensing, we consider two scenarios of practical interest: 1) a single sensor case in which only one spectrum band is observed at one time, and 2) a multiple sensor case in which multiple spectrum bands are observed simultaneously. For each case, scenarios with and without a scanning delay constraint are investigated. Using mathematical tools from optimal stopping theory, we develop optimal spectrum scanning algorithms to minimize a cost function that strikes a desirable tradeoff between detection performance and sensing delay. For the non delay-constrained scenario, we show that the optimal scanning algorithm is a concatenated sequential probability ratio test (C-SPRT). For the delay-constrained scenario, we show that the optimal scanning algorithms involve a high implementation complexity. Thus, we propose multiple low-complexity truncation schemes with C-SPRT sensing as an alternative to the high-complexity delay-constrained sensing algorithms. Simulation results show significant improvement in sensing time for our algorithms in comparison to the sensing procedures with a fixed sample-size. Besides spectrum sensing, optimal channel estimation and probing can enhance the throughput performance of CR. For instance, the CR user does not only look for a free channel, but also a good one. We propose a novel sequential channel estimation approach for multiband CR systems. We introduce a general model and test two scenarios of practical interest. The two scenarios are: 1) CR users optimally estimate all the available bands, and 2) CR users find one good channel with a large gain. In particular, we use a sequential search in which the CR users estimate the available channels one by one. During the search, the CR users determine whether to terminate the current channel estimation process and switch to the next channel based on the training symbols received so far. Our objective is to design a switch function, an estimator, and a stopping rule that minimize a combination of estimation time and error. For the multiband estimation scenario, we show that the optimal rule is to find the optimal number of symbols required for each channel in a joint optimization problem. For the good channel search problem, we show that the optimal decision rules that minimize a properly chosen cost function have a simple structure. In particular, both the termination and switching rules are threshold based. Simulation results demonstrate the effectiveness of our channel estimation and search strategies in comparison to the traditional approaches. Real-time implementation and algorithm benchmarking is the final stage to transform the CR technology from research labs to the industry. In the final phase of this research, we introduce the necessary tools to implement spectrum sensing algorithms and provide a case study. Specifically, we develop a large scale spectrum sensing framework for cognitive radio systems. We first design a framework that is capable of collecting and combining sensing results from multiple nodes equipped with software defined radios. We then implement and compare the performance of three sensing algorithms in standalone nodes equipped with a Universal Software Radio Peripheral (USRP). In particular, the three algorithms are: 1) a multiband time-based energy detector, 2) a multiband frequency-domain-based energy detector, and 3) a low complexity sequential multiband spectrum sensing algorithm with a delay constraint. Furthermore, we develop a simple and efficient procedure for choosing threshold values and the number of samples required for the detection process. The procedure is based on observational statistical averages. Finally, we use the developed framework for sensing large bandwidth using multiple USRP nodes and ORBIT testbed. Spectrum sensing results are provided to demonstrate the efficiency of the developed algorithms and framework for sensing large spectrum bandwidth in a real-time environment.

Download

Share

COinS