Resource allocation and finite block length regime

Abstract

This thesis aims to introduce enhanced versions of the Temporal Compressive Resource Allocation technique, which is used for resource allocation (bits and transmission power) in multi-carrier data communication systems for application in Power Line Communication and to discuss the finite block length regime in single-carrier digital data communication systems when the receiver is unaware of the transmission power and may be unaware of the communication channel. In this sense, the resource allocation problem is first formulated and a review of the original technique is shown, highlighting the favorable points of the technique and the possible suggested enhancements. Thus, the proposed enhancements are accomplished through the introduction of two changes that increase the effectiveness of the resource allocation process when the maximization of the data rate is desired. The first change refers to the use of the minimum signal-to-noise ratio of each set of microslots to avoid peaks in the symbol error rate in some microslots associated with the mains frequency in electrical power networks. The second change imposes the use of the normalized signal-to-noise ratio matrix to estimate the correlation between microslots since it is considered a more reliable parameter to provide this type of information because it is the result of less processing. Numerical results, based on a comparison between the partial enhanced version, a complete enhanced version, and the original technique, show that both enhanced versions are able to guarantee the symbol error rate upper bound with reduced data rate penalty compared to the original technique. In addition, it is possible to notice that the use of the normalized signal-to-noise ratio matrix results in a greater reduction in computational complexity. In the sequel, the problem of data transmission through blocks of finite length is formulated in single-carrier digital data communication systems in a scenario in which the receiver must estimate the transmission power and the additive noise power considering blocks of symbols of finite length, which generates randomness in the calculation of the signal-to-noise ratio at the receiver. To circumvent this problem, it is proposed the use an extra gap factor to minimize the impact of such randomness and to statistically evaluate the consequences of using blocks of finite length by modeling a random variable when considering the 푀-QAM and 푀-PAM constellations. Numerical analyses show that when the receiver is aware of the communication channel, high values of the extra gap factor are necessary to ensure reliable data communication. On the other hand, when the communication channel is unknown by the receiver, blocks of greater lengths or greater values of the extra gap factor are crucial to ensure the reliability of data communication, especially in the presence of a malicious device, which is capable of overhearing the signal transmitted between two legitimate nodes that belong to the data network.