Search

With the latest technology of vectoring, DSL data rates in the order of 100Mbps have become a reality that is under field deployment. The key is to cancel crosstalk from other lines, which is also known as multiuser MIMO cancellation for wireless communications. During the DSL system upgrade phase of field deployment, mix of legacy and vectoring-enabled VDSL lines is inevitable and a channel estimation solution for the entire mix is needed before vectoring can be enforced. This paper describes a practical method for crosstalk channel estimation for downstream vectoring, assuming that a vectoring-enabled DSLAM forces DMT symbol-level timing to be aligned for all of the lines, but also assuming that the location of synch symbols are aligned only among vectoring-enabled lines. Each vectoring-enabled receiver is capable of reporting error samples to vectoring-DSLAM. The estimation method is not only practical, but also matches the performance of Maximum-Likelihood estimator for the selected training sequences., Comment: 4 pages, 3 figures

This paper is concerned with the fading MIMO-MAC with multiple receive antennas at the base station (BS) and multiple transmit antennas at each mobile terminal (MT). Two multiple-access techniques are considered for scheduling transmissions from each MT to the BS at the same frequency, which are space-division multiple-access (SDMA) and time-division multiple-access (TDMA). For SDMA, all MTs transmit simultaneously to the BS and their individual signals are resolved at the BS via multiple receive antennas while for TDMA, each MT transmits independently to the BS during mutually orthogonal time slots. It is assumed that the channel-state information (CSI) of the fading channel from each MT to the BS is unknown at each MT transmitter, but is perfectly known at the BS receiver. Thereby, the BS can acquire the long-term channel-distribution information (CDI) for each MT. This paper extends the well-known transmit-covariance feedback scheme for the point-to-point fading MIMO channel to the fading MIMO-MAC, whereby the BS jointly optimizes the transmit signal covariance matrices for all MTs based on their CDI, and then sends each transmit covariance matrix back to the corresponding MT via a feedback channel. The main goal of this paper is to characterize the so-called multiuser power region under the multiuser transmit-covariance feedback scheme for both SDMA and TDMA. The power region is defined as the constitution of all user transmit power-tuples that can achieve reliable transmissions for a given set of user target rates. Simulation results show that SDMA can achieve substantial power savings over TDMA for the fading MIMO-MAC. Moreover, this paper demonstrates the usefulness of the multiuser power region for maintaining proportionally-fair power consumption among the MTs.