Abstract
A few years ago the communications industry announced a plan for the introduction of fifth-generation (5G) wireless network. Unlike the previous successive generations which are a slight iteration of their predecessor technologies, 5G is going to be a major leap in the communication industry by introducing several new applications, which have never been covered by any telecommunication standards. Example of such applications are the Ultra-reliable low-latency communications, vehicle-to-everything and the internet of things. And in order to meet these new challenges, researchers world over are seeking new methods and techniques that will make the 5G revolution a reality. Index modulation for multicarrier systems which is the focus of this thesis is one of the new physical layer technologies that hold the promise of improving data rates without increasing the system bandwidth, as well as improves energy efficiency. Orthogonal frequency division multiplexing with index modulation, for example, conveys information by selecting and activating a subset of OFDM subcarriers for M-ary symbol transmission. In addition to the M-ary symbols, information is encoded into the indices of the active carriers to improve spectral efficiency. The energy saved from the inactive subcarriers can be redistributed through the power redistribution policy (PRP) among the active subcarriers to improve SNR or can be saved, by implementing the power saving policy (PSP), therefore improving energy efficiency. In order to provide a valuable reference and guideline for the design of OFDM-IM systems, this thesis has analyzed the error performance of both single active and multi-active subcarriers per cluster systems and as a result, provides a generalized bit error rate (BER) expression for OFDM-IM in the AWGN channels. Using (n=2, k=1, M-ary) OFDM-IM configuration, a closed-form BER expression of OFDM-IM in a Rayleigh fading channel has also been derived. Analytical results show that the BER of an OFDM-IM system is a weighted sum of exponentials and Q-functions which conforms well with the numerical results up to a BER of 10