From Information Rate Computation to Communication System Design

Abstract

Abstract-This paper investigates the application of information-theoretic techniques to the design of advanced communication and storage systems. We use the information rate (IR), evaluated using a recently proposed efficient numerical technique, as the key parameter characterizing the system performance. As case studies, we consider three different scenarios: (i) single-carrier (SC) and multi-carrier (MC) transmission systems over linear dispersive channels, (ii) digital storage systems impaired by media noise, and (iii) optical communication systems with direct photo-detection. The obtained results show that a clever IR analysis can provide significant insights into system performance and design in several scenarios of interest in current applications.

The Contribution of the Unresolved Extragalactic Radio Sources to the Brightness Temperature of the Sky

“…The low-flux tail of the model compared with the experimental data and our fit Q(S) are shown in Figure 2. This evolution model is able to predict accurately the slope of the low-flux tail and we used it to get the value of ε 2 , which is independent on the frequency (Franceschini 2007). The values of ε 2 , obtained at 1.4 and 5 GHz, are fully compatible with the average value of ε 1 previously evaluated.…”

On information theoretic aspects of single- and multi-carrier communications

“…3 It can be easily shown that MC and SC schemes yield the same IR in the case of memoryless AWGN channel [8]. As a consequence, the IRs I(x k ; ξx k + w k ) and I(x k ; ρx k + w k ) in (7) and (8), respectively, can also be interpreted as the IRs of two memoryless AWGN channels with MC input.…”

“…Typically, bit-loading algorithms allocate in each channel the highest possible number of bits that guarantees a given operational bit error rate (BER). In order to make a fair comparison between uniform input power SC and MC schemes with water-filling, however, we use a bit-loading algorithm (first introduced in [8]) which optimally allocates a given fixed number of bits to be transmitted by maximizing the overall system IR. 5 This allows to fix the signal space cardinality, thus limiting the system spectral efficiency.…”

“…We remark that the same SNR definition was used in[8], although it was erroneously stated that SNR was the ratio between the average received sample energy and N 0 . The two definitions, i.e., referring to average received and transmitted sample enery, are equivalent in the case of normalized channel impulse response energy and uniform transmitted power spectrum.…”