Matteo Campanella

Optimum bandwidth-distance performance in partial response CPM systems

The problem of optimizing the baseband pulse shape in order to obtain the minimum effective bandwidth for a prescribed value of minimum Euclidean distance is treated for partial response CPM. Some comparisons with standard pulses are performed and the results are presented in the form of bandwidth-distance and spectral density plots.

On the channel capacity for constant envelope signals with effective bandwidth constraint

The problem of the optimization of the information rate with an assigned cost function is addressed. In the case of a constant envelope channel with a prescribed effective bandwidth as a cost function, an expression of the channel capacity is derived in the presence of Gaussian noise for large values of signal-to-noise ratio. The analysis of the optimum source statistics has also shown that the source signal is Gaussian, and therefore its autocorrelation function and power spectral density have been determined in terms of the required channel capacity. >

A design technique for spectral shaping in CPM systems

We describe a technique of pulse shaping able to control the power spectral density of a full-response CPM signal in various frequency ranges, under the constraint of a prescribed minimum Euclidean distance. Some examples illustrating the potentialities of the method and some comparisons with standard pulses are presented.

Effective bandwidth optimization for noncoherent full-response CPM systems

The optimization of the baseband pulse shape for full response CPM signals, when a noncoherent lowest complexity receiver is used, is carried out. The effective bandwidth is chosen as the objective of the optimization. In order to obtain the minimum error probability for high signal-to-noise ratio, orthogonal signaling is taken into consideration. A closed form for the optimum pulse shape, the value of the optimum modulation index and of the optimum effective bandwidth is determined. The power spectral density and the fractional out-of-band power have been compared with some phase functions considered in the literature. >

Statistical analysis of RaptorQ failure probability applied to a data recovery software

In this work, we have implemented a data recovery software integrating the most recent rateless codes, i.e., RaptorQ codes. Thanks to the above-mentioned software, it is possible to recover data loss occurring on several kinds of network conditions. We have performed a statistical analysis of failure probabilities at several configurations of RaptorQ parameters. We have found a good agreement with the theoretical values of a random linear fountain code over Galois Field GF(256). Moreover, we have shown that the probability of having a certain number of failed decoded source blocks - when sending a fixed-size file - follows a Poisson distribution.

Fading mitigation coding techniques for space to ground free space optical communications

In this manuscript, a Geostationary satellite-to-ground Free Space Optics (FSO) downlink channel model has been implemented, which is able to predict temporal irradiance fluctuations caused by scintillation at a wide range of turbulence conditions and for different values of the zenith angle. In order to mitigate fading events that occur in FSO communications, we have also tested the performance of three different families of Rateless Codes (Luby Transform, Raptor and RaptorQ) into our model and found that RaptorQ is the best candidate to mitigate errors in FSO links.

On the ∗-cocharacter sequence of 3×3 matrices

Abstract Let M 3 (F) be the algebra of 3×3 matrices with involution * over a field F of characteristic zero. We study the ∗ -polynomial identities of M 3 (F) , where ∗=t is the transpose involution, through the representation theory of the hyperoctahedral group B n . After decomposing the space of multilinear ∗ -polynomial identities of degree n under the B n -action, we determine which irreducible B n -modules appear with non-zero multiplicity. In symbols, we write the nth ∗ -cocharacter χ n (M 3 (F),*)=∑ r=0 n ∑ λ⊢r,h(λ)⩽6 μ⊢n−r,h(μ)⩽3 m λ,μ χ λ,μ , where λ and μ are partitions of r and n−r , respectively, χ λ,μ is the irreducible B n -character associated to the pair (λ,μ) and m λ,μ ⩾0 is the corresponding multiplicity. We prove that, for any n, the multiplicities m λ,μ are always non-zero except the trivial case λ=(1 6 ) and μ=∅ .

Design of efficient codes for the AWGN channel based on decomposable binary lattices

This work is concerned with the use of binary decomposable lattice codes over the QAM Gaussian channel. First, we investigate the structure of such class of lattices: we derive consistency conditions for the binary codes appearing in their decomposition and express their nominal coding gain and some bounds for their error coefficient in terms of the parameters of the component codes. Then we describe a general multistage bounded‐distance decoding algorithm with low complexity and we evaluate its performance. Finally, we develop a design example and report the corresponding simulation results; as a reference some comparisons with standard TCM codes are also presented.

A study on spectral optimisation in partial response CPM signals

An optimisation technique for the minimisation of the effective bandwidth in partial response CPM signals is described. The minimum Euclidean distance is used as optimisation constraint because the bit error probability is a function of this parameter for high values of signal to noise ratio. The procedure has been implemented for a correlation length corresponding to two signalling intervals. The optimisation problem leads to a system of two nonlinear differential equations for the pulse shape, together with some boundary conditions. Numerical solution of the differential equations has been performed; initial conditions have been adjusted, to satisfy the boundary conditions using an iteration method based on the Newton-Raphson algorithm. The whole procedure is based on a particular assumption about the minimising difference sequence for the minimum Euclidean distance. This assumption has been tested and the limit value of the modulation index satisfying the test is also reported.

A simple joint estimation-detection technique for OFDM systems

In this work a simple approach for the joint estimation-detection in a frequency selective severe fading environment of OFDM signals adopting PSK constellations is presented. A linear predictor of suitable order is adopted for the channel estimation in the frequency domain. The predictor coefficients are estimated on the basis of a sample estimation of the autocorrelation of the channel frequency response, aided by a preliminary differential decoding, in a blockwise manner. The detection technique proposed here is based on a simple tree search where a small number of best survivor paths are maintained at each step. Despite the simplicity of above detection approach, the simulation results show its effectiveness when the number of carriers is sufficiently high. Appealing features of this method are that it does not rely on any detailed model of the fading channel and that the channel coefficients are allowed to change independently in correspondence of different blocks of OFDM symbols.