Compressive Identification of Active OFDM Subcarriers in Presence of Timing Offset

TL;DR

This paper presents a compressive sensing framework for identifying active OFDM subcarriers in wideband signals, accounting for timing offsets, which is crucial for spectrum sensing in cognitive radio systems.

Contribution

It introduces a joint dictionary learning and sparse recovery method to identify active subcarriers under unknown timing offsets in sub-Nyquist sampled OFDM signals.

Findings

Reliable active subcarrier identification achieved with the proposed method.

Effective joint dictionary learning for timing offset estimation.

Framework applicable to wideband spectrum sensing in cognitive radios.

Abstract

In this paper we study the problem of identifying active subcarriers in an OFDM signal from compressive measurements sampled at sub-Nyquist rate. The problem is of importance in Cognitive Radio systems when secondary users (SUs) are looking for available spectrum opportunities to communicate over them while sensing at Nyquist rate sampling can be costly or even impractical in case of very wide bandwidth. We first study the effect of timing offset and derive the necessary and sufficient conditions for signal recovery in the oracle-assisted case when the true active sub-carriers are assumed known. Then we propose an Orthogonal Matching Pursuit (OMP)-based joint sparse recovery method for identifying active subcarriers when the timing offset is known. Finally we extend the problem to the case of unknown timing offset and develop a joint dictionary learning and sparse approximation algorithm, where in the dictionary learning phase the timing offset is estimated and in the sparse approximation phase active subcarriers are identified. The obtained results demonstrate that active subcarrier identification can be carried out reliably, by using the developed framework.