Search

Bit rate fairness is an important concern in current G.fast and the forthcoming fifth generation DSL systems where the frequency is extended to hundreds of MHz and crosstalk couplings reach unprecedented levels. In this paper, we study the price of fairness (PoF) in dynamic spectrum management (DSM) enabled DSL systems. We propose optimal low complexity PoF-constrained max-min fairness (MMF), weighted max-min fairness (WMMF), proportional fairness (PF), and (p,α)-PF algorithms. The proposed algorithms inherently provide weight factors which can be used to reduce the computational complexity of user encoding or decoding order optimizing algorithms used in nonlinear vectoring. Our simulation results show that although PoF grows exponentially with the minimum bit rate, fairness can be improved considerably with a relatively small price in DSM level 2 spectrum balancing and particularly in DSM level 3 using minimum mean square error (MMSE) generalized decision feedback equalizer (GDFE). It is also seen that the proposed optimal PoF-constrained MMF algorithm can reach the solutions of PF and (p,α)-PF measures for some PoF. That is, the PoF-constrained MMF or WMMF algorithms can be used instead of the non-linear PF and (p,α)-PF measures, which often result in computationally intensive solutions. [ABSTRACT FROM AUTHOR]

A cyclostationary feature-based wideband spectrum sensing is proposed in which cyclic frequency offset is considered. The received signal passes through a rough and flexible filter which its effective band tuned to a specific part of the received signal spectrum. It belongs to a target signal, which possibly exists in the received signal. Some cyclic frequencies of the target are employed to derive the normalized least mean square algorithm that estimates the filter output from its frequency shifted samples. Based on the weight estimate’s norm, the algorithm is tuned to find the accurate cyclic frequencies. If it succeeded, it means that the target signal is present. On the other hand, a complete search without major weights estimate means that the target signal is absent. In addition, the overall system, named cyclic frequency lock loop can be reformed as a demodulator for sinusoidal frequency modulation signals. The method is very easy to implement and has less computational complexity comparing with its other spectrum sensing counterparts. Simulation results validate the efficiency of the proposed method. [ABSTRACT FROM PUBLISHER]

Recently, the problem of optimal spectral balancing (OSB) for digital subscriber lines (DSL) with constrained transmit power has been solved using Lagrange's dual optimization technique and a weighted sum rate maximization (WSRM) approach. In many cases, the total power constraint is not binding. Although, this means a huge computational complexity reduction, the algorithm fails to reach certain points on the rate region (RR). In this paper, an in-depth analytical view of the WSRM approach is provided, and it is shown that when the RR is not strictly convex, the WSRM approach fails to reach certain points on the RR. Moreover, using N-dimensional geometry, a novel iterative facet dividing algorithm (IFDA) capable of reaching any point on the RR is proposed. Analytical and simulation results show that our technique is much more reliable and considerably faster than current algorithms. Moreover, it can be used for a wide range of problems which use WSRM approach, including OSB in the general case. [ABSTRACT FROM AUTHOR]

Dynamic spectrum management (DSM) refers to a wide range of techniques for counteracting crosstalk in digital subscriber line (DSL) networks. DSM is categorized into three levels based on the degree of coordination among users. In this article, we investigate optimal joint level 2 and 3 DSM for upstream DSL. We will discuss the difficulties of finding the universally optimal solution and we propose an optimal algorithm, referred to as IF/MAC-OSB, under some practical and implementation assumptions for this problem. Using computer simulations, we show that IF/MAC-OSB is capable of increasing the user bit rates considerably compared to several other DSM techniques. The proposed algorithm involves using the minimum mean squared error (MMSE)-generalized decision feedback equalizer (GDFE) together with Lagrange dual optimization. We address several aspects of the problem including the optimal decoding order in the GDFE receiver, GDFE error propagation, and the computational complexity of the algorithm. We also study effects of channel model randomness and upstream power back-off utilization on the performance of the algorithm. [ABSTRACT FROM AUTHOR]