In this paper, we address the main design issues and performance characteristics of an echo canceller for digital DVB-T/H on-channel repeaters (OCRs). The low-complexity channel cancellation technique here described is based on the introduction of a negative feedback loop capable to counterbalance the coupling effects between the OCR antennas (echoes). In the initial phase, locally generated pulse trains are transmitted to estimate the coupling channel between the transmitting and the receiving antennas. This first echo estimation is used to set-up the cancelling unit, whose purpose is to generate a properly filtered replica of the repeated DVB-T/H signal compensating the echoes. After the initial set-up of the system, efficient tracking of the echoes can be performed through traditional least mean square (LMS) techniques. After developing a proper theoretical model, we introduce and discuss some important performance figures such as mean rejection ratio (MRR) and echo suppression at nominal position (ESNP). Then, a detailed description of our implementation of the echo canceller on a field programmable gate array (FPGA) board is reported. Finally, the performance of our prototype are evaluated and the measurement results are compared with those derived by our analytical model.

We analyze a two-stage echo cancellation technique for on-channel repeaters (OCRs) based on the reconstruction of the coupling channel impulse response via digital filtering. The estimation of the coupling channel between the transmitting and the receiving antennas is achieved through the injection in the repeated signal of locally generated low-power pseudonoise sequences or trains of short pulses. The analytical model developed in this paper permits the investigation of some important performance figures, such as the mean rejection ratio, the echo suppression and the echo suppression at nominal position. In addition, a new analytical framework dealing with canceller stability issues, arising due to the non perfect echo cancellation, is derived. Specifically, an upper bound on the probability of instability of the canceller is provided. As shown in the numerical results, this framework permits the identification of the main OCR design guidelines. Finally, we will validate our analytical models by showing their accordance with the experimental measurements realized with the OCR prototypes we developed.

In this paper we provide the outcomes of a measurement campaign aimed at characterizing the UMTS femtocell propagation in an indoor scenario. The accuracy of a commonly adopted estimator of the signal-to-noise ratio is discussed, then the minimum measurement interval needed to derive an accurate characterization of the average received power is discussed. Finally we compare the measurement outcomes with those predicted using ray tracing tools and analytical models, showing the achievable accuracy levels.