Distributed Joint Source Coding and Trellis Coded Modulation for Symbol-Based Markov Sources

Abstract

A distributed joint source-channel coding scheme based on a unity-rate code (URC)-assisted trellis coded modulation (TCM) is proposed, which exploits the spatio-temporal correlation of symbol-based sources. More specifically, asymmetric distributed source coding of two spatially correlated Markov sources is considered, where one of the sources is assumed to be perfectly decoded and to be available at the receiver of the other source as side information. In order to exploit the temporal correlation statistics, an iterative decoding process exchanging extrinsic information between the amalgamated URC-assisted TCM and a soft-symbol source decoder employing a modified symbol-based maximum a posteriori algorithm is invoked. Furthermore, the Slepian-Wolf (SW) bound of symbol-based sources having spatiotemporal correlation is derived and the benefits of exploiting the spatio-temporal correlation using the proposed coding scheme are demonstrated by our extrinsic information transfer chart analysis. It is shown from our simulation results that upon exploiting the spatio-temporal correlation of the sources, the proposed coding scheme is capable of operating within 0.02 bit of the SW bound.

Joint Decoding and Estimation of Spatio-Temporally Correlated Binary Sources

“…BER performance of the proposed scheme estimating both the spatial and temporal correlation parameters, spatial correlation parameter only, and without estimating any correlation parameter when there is no knowledge (NK) of the correlation parameters at the receiver as compared to the ideal system with perfect knowledge (PK) of the correlation parameters at the receiver after 20 iterations for sources having p = q = 0.6 and p = q = 0.8. The vertical lines indicate the associated capacities [5], [10].…”

“…Hence, the authors of [8]- [10] exploited the spatio-temporal correlation of the sources and demonstrated that further performance improvements can be achieved by exploiting both the spatial and temporal correlations, as compared to spatial correlation only.…”

“…In order to avoid this problem, it was assumed in [8]- [10] that the partial or full statistical knowledge of the spatiotemporal correlation is available at the receiver. However, this idealistic assumption can only be justified by relying on additional resources to forward the correlation coefficients to the receiver.…”

“…We employ a bandwidth-efficient DJSCC scheme constituted by a serialconcatenated trellis coded modulation (TCM) scheme [12] and a non-binary unity-rate code (URC) [13]. The serial concatenated TCM scheme and an URC can achieve better performance while having lower computational complexity than the TTCM-based DJSCC scheme [10]. The symbol-based maximum a posteriori (MAP) algorithm [14] is modified for TCM decoding in order to exploit the spatio-temporal source correlation, while performing iterative decoding.…”

Distributed Joint Source-Channel Coding-Based Adaptive Dynamic Network Coding

“…Additionally, the attainable performance gain when activating such decoding iteration would be limited in the block fading channel environment [17], as the channel coding itself cannot considerably improve the Bit Error Ratio (BER) [18]. Recently, two spatially correlated Markov sources based transmission was supported using DJSC scheme in [19]. Additionally, a novel lattice-based coding algorithm was proposed in [20], which shows a robust transmission over noisy channel.…”