Channel modeling is critical for evaluating various protocols designed for different network layers, as credibility of the corresponding evaluation largely depends on the accuracy of the propagation models. Dynamic vehicular network topology and propagation environment define the communication channel behavior. In this paper, we study the received signal strength (RSS) based channel characterization dataset collected from an urban intersection scenario for Dedicated Short Range Communication (DSRC) in the 5.9 GHz band. We conducted a large-scale RSS measurement campaign in one of the densest traffic environments in the US. The channel characterization data was collected using ten DSRC-enabled vehicles in an Orange County, CA intersection environment. Using the dataset, we study and identify the key aspects of the network topology and physical environment that collectively describe the channel behavior and determine realistic wireless propagation channel models for urban intersections. In particular, we show that the path losses are harsh between vehicles on perpendicular roads. Our results may prove to be instrumental in evaluating vehicleto- vehicle (V2V) safety applications.

The LTE standards account for the use of relays to enhance coverage near the cell edge. In a traditional topology, a mobile can either establish a direct link to the base station (BS) or a link to the relay, but not both. In this paper, we consider the benefit of multipoint connectivity in allowing user equipment (UEs) to split their transmit power over simultaneous links to the BS and the relay, in effect transmitting two parallel flows. We model decisions by the UEs as to: (i) which point of access to attach to (either a relay or a relay and the BS or only the BS); and (ii) how to allocate transmit power over these links so as to maximize their total rate. We show that this flexibility in the selection of points of access leads to substantial network capacity increase against when nodes operate in a fixed network topology. Individual adaptations by UEs, in terms of both point of access and transmit power, are interdependent due to interference and to the possibility of over-loading of the backhaul links. We show that these decisions can converge without any explicit cooperation and derive a closed-form expression for the transmit power levels.

Long Term Evolution (LTE) release 12 proposes the use of dual connectivity in heterogeneous cellular networks, where a user equipment (UE) maintains parallel connections to a macrocell base station and to a low-tier node such as a picocell base station or relay. In this paper, we propose distributed multi-objective power control where each UE independently adapts its transmit power on its dual connections, where the uplinks could be of unequal bandwidth and have non-ideal backhaul links. In the proposed optimization, the UEs dynamically switch between data rate maximization and transmit power minimization as the backhaul load varies. To address the coupling between interference and the backhaul load, we propose a low-overhead convergence mechanism which does not require explicit coordination between UEs and also derive a closed-form expression of the transmit power levels at equilibrium. Simulation results show that our scheme achieves higher aggregate end-to-end data rate and significant power saving in comparison to a scheme that employs a greedy algorithm and a scheme that employs only waterfilling.

LTE-Advanced standards contemplate support for heterogeneous networks, through the use of relays, as well as the ability to perform carrier aggregation in order to achieve higher data rates. In this paper, we model the User Equipment's (UE) ability to select its current point-of-access to the network (either a relay or the base station), aggregate additional channels, and perform power adaptations. The adaptations by UEs affect one another, by impacting the interference conditions in the network as well as congestion conditions in the relay-to-base station links. We propose an adaptation mechanism whereby the UEs consider these factors and select their transmit power, channels, and point of access to the network to optimize a combination of achieved data rate and energy consumption. Our results show these adaptations to be more efficient than adaptations that simply consider current channel quality indicators, combined with greedy channel aggregation.

Multi-channel operation in an ad hoc network can improve robustness and reliability by efficiently managing interference and reducing contention. In this paper, we model four dynamic channel assignment techniques under the same set of assumptions, comparing the efficiency of their power and channel allocations. As the number of channels increases, the differences in the performance of the four techniques become more pronounced. Among the techniques studied, the conflict graph-based technique achieves the highest number of feasible links and the lowest average power consumption.