An Alternative to Decoding Interference or Treating Interference as Gaussian Noise

Abstract

This paper addresses the following question regarding Gaussian networks: Is there an alternative to decoding interference or treating interference as Gaussian noise? To state our result, we study a decentralized network of one primary user (PU) and one secondary user (SU) modeled by a two-user Gaussian interference channel. In one scenario, the primary transmitter is constellation-based and PUs codebook is constructed over its modulation signal set. Assuming SU is aware of the constellation points of PU, the interference plus noise at the secondary receiver is modeled by a mixed Gaussian process. We show that SU can achieve larger rates by matching its decoder to the actual interference plus noise compared with the case where the secondary receiver performs nearest neighbor decoding (NND). In another scenario, we assume that PU utilizes a predetermined point-to-point code. We ask if SU can utilize its knowledge about PUs codebook without decoding PUs codewords. The proposed strategy assumes each transmitted codeword of SU overlaps with infinitely many transmitted codewords of PU, referred to as the unequal codeword-length (UCL) strategy. The secondary receiver views PU as a virtual user that is constellation-based and its modulation signal set is the codebook of the actual PU. UCL is compared with other strategies, namely, interference cancellation (IC), joint decoding, and NND. It is shown that UCL can outperform both IC and NND simultaneously.Index Terms-Cognitive radio, constellation-based transmission, decentralized networks, mixed Gaussian interference, unequal codeword-length.

A View of Information-Estimation Relations in Gaussian Networks

“…Another reason is that discrete inputs often outperform Gaussian inputs in competitive multi-user scenarios, such as the interference channel, as will be demonstrated in Section 7. For other examples of discrete inputs being useful in multi-user settings, the interested readers is referred to [43][44][45][46].…”

“…An open question, for n = 1, is what value of p provide the smallest gap and whether it coincides with the ultimate "shaping loss". For the AWGN channel there exist a number of other bounds that use discrete inputs as well (see [46,[56][57][58] and references therein). The advantage of using Ozarow-Wyner type bounds, however, lies in their simplicity as they only depend on the number of signal constellation points and the minimum distance of the constellation.…”

“…This is enabled by using non-Gaussian inputs and designing the decoders that treat interference as noise by taking into account the correct (non-Gaussian) distribution of the interference. Such scenarios were considered in [44,46,49], and shown to be optimal to within either a constant or a O(log log(snr)) gap for all regimes in [45].…”

“…We also note that in [47,70] it was conjectured that the optimal input for C 1 (snr, snr 0 , β) is discrete. For other recent works on optimizing the TIN region in (51), we refer the reader to [43,46,49,78,79] and the references therein.…”

An Information-Spectrum Approach to the Capacity Region of the Interference Channel

“…(a) For t = 1, 2, ..., n, compute the values of P(y 1,t |y t−1 1 ) and α t (s t ) recursively according to Equations (19) and (20).…”