A Spectral Compressive Resource Allocation Technique for PLC Systems

Statistical analysis and modeling of a novel parameter for resource allocation in multicarrier PLC systems

“…In this regard, the nSNR coherence bandwidth as used in Colen et al is the most appropriate parameter for exploiting the existent relationship among adjacent subchannels, since it precisely reflects the flatness of the nSNR. Grounded on the definition of the coherence bandwidth, the nSNR coherence bandwidth is defined as follows:…”

“…For this case, the choice of α =0.9 results in B c , H =5.1 MHz and MHz. In this specific situation, the use of B c , H in resource allocation technique may introduce remarkable data rate loss, since subcarriers with rather disparate nSNR values are grouped within the same set, which represents a chunk, see Colen et al for more details. For instance, if the bandwidth of a subchannel in a PLC channel is Δ B =50 kHz, then will group 40 subsequent subcarriers, while B c , H will group 102 subsequent subcarriers.…”

“…Note that Figure considers 2 different values of α to evaluate both coherence bandwidths. The first one is α =0.9, which yields a computational complexity reduction of 92.4% for and 93.4% for B c , H if we use the SCRA technique . The second one is α =0.5, which yields a computational complexity reduction of 98.6% for and 98.9% for B c , H , once again using the SCRA technique.…”

“…Optimal resource allocation, as discussed in Tunc et al, may be an impracticable aim, since it demands rather high computational complexity and signaling overhead. On the other hand, proposed the spectral compressive resource allocation (SCRA) technique, which reduces computational complexity and signaling overhead associated with resource allocation by grouping subcarriers with similar normalized signal‐to‐noise ratio (nSNR) into chunks. Moreover, a closer look at practical resource allocation techniques for OFDM‐based PLC systems and references therein shows that most of them makes use of the nSNR, which is the signal‐to‐noise ratio (SNR) when the transmitter allocates unit transmission power to each subchannel.…”

“…In this context, we focus on a comprehensive disscussion on the so‐called nSNR coherence bandwidth, which can be a very interesting parameter to reduce computational complexity of resource allocation techniques applied to OFDM‐based PLC systems. Due to its characteristics, the nSNR coherence bandwidth is capable of precisely characterizing the flatness of the nSNR and therefore is more suitable than the coherence bandwidth for determining the chunk length/bandwidth in resource allocation techniques, such as in Colen et al and Colen et al Moreover, considering a data set constituted by PLC channel estimates and additive noises measurements of in‐home, low‐voltage outdoor, and hybrid PLC‐wireless environments, we statistically model the nSNR coherence bandwidth. Such modeling is useful to conduct a performance analysis of the influence of chunk length/bandwidth on resource allocation techniques applied to OFDM‐based PLC systems.…”