Search

©2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., Orthogonal Frequency Division Multiplexing (OFDM)-based networks rely on time synchronization to obtain their best performance. Time synchronization with neighboring nodes can be satisfied by increasing the cyclic prefix (CP) length (at the expenses of spectral efficiency reduction) We show that by optimizing the transmit pre-compensation and receive post-compensation we can meet the time synchronization constraints and keep the CP at its minimum value. This concept is applied to paired-bands Frequency Division Duplexing (FDD) systems which tend to show inefficient occupancy of the uplink (UL)-band due to the traffic asymmetry. We consider the possibility of deploying multiple Time Division Duplexing (TDD) small eNBs (SeNBs) in the unused UL spectrum. In this scenario, time synchronization with macro eNB (MeNB) and neighboring SeNBs becomes essential. Two algorithms are proposed in order to ensure orthogonality of OFDM transmissions network-wide., Peer Reviewed, Postprint (author's final draft)

©2016 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works., This paper presents long-term resource provisioning schemes for dense multi-cell OFDMA-based networks, where multiple cells with possibly overlapping coverage areas compete for the same set of resources. The optimization is done over the long term, being independent of the specific users connected to each cell but dynamic enough to follow significant variations of the traffic load. In this sense, we focus on the average resource utilization (RU) and propose schemes to minimize the maximum RU of all cells. Firstly, we consider orthogonal resource usage among cells. Optimal closed-form expressions for the long-term resource provisioning are derived for this case. Secondly, we assume that resources can be reused at non-overlapping cells. In this case, the resource provisioning is solved in two steps: i) the number of resources required per cell is obtained by discretizing the optimal solution of a convex problem, and then ii) the specific resources to be utilized by each cell are determined by using graph coloring. In contrast to previous works, graph coloring can be applied to get an implementable solution for any condition of the cell loads. Simulation results show a significant reduction of the maximum RU, which translates into an increase of the served traffic and a reduction of the packet delay, as compared to static resource provisioning schemes., Peer Reviewed, Postprint (published version)

This paper proposes a joint user scheduling and transmit direction selection procedure for dynamic TDD in 5G dense small cell networks, where the transmit direction (i.e. downlink (DL) or uplink (UL)) selected per small cell is dynamically optimized together with the user scheduling at every frame. We focus on the maximization of a general utility function that takes into account the DL/UL traffic asymmetries of each user and the interference conditions in the network. After relaxation of the integer variables that subsume the user scheduling and transmit direction selection, the problem is decomposed and efficiently solved in parallel at every frame. Simulation results show significant gains in DL and UL average rates for different traffic asymmetries, network densities, and user densities as compared to existing schemes for dynamic TDD., Peer Reviewed, Postprint (published version)

The optimal transmission scheme for a multiple-input multiple-output point-to-point channel (MIMO-P2P) is dependent on the type of received interference, which can be either proper or improper. We will show that the presence of improper interference is beneficial in terms of improving the achievable rate compared to a case with proper interference. These benefits are due to using widely linear precoder and receiver. Additionally, we investigate a cellular scenario where two types of users are coexisting, ones that are receiving proper interference while others which might be receiving improper interference, trying to devise the best transmission scheme to be adopted. We propose an improper Gaussian signaling-based scheme that allows controlling system performance and fairness in terms of bitrate through a single parameter., Peer Reviewed, Postprint (published version)