Delay–Throughput Enhancement in Wireless Networks With Multipath Routing and Channel Coding

Multipath routing and adaptive channel coding are two well-known approaches that have been separately applied to wireless networks to improve the effective throughput. However, it is usually expected that achieving a high throughput would be at a noticeable cost of increasing the average end-to-end delay and causes major degradation in the overall network performance. In this paper, we show that a combination of multipath routing and adaptive channel coding can improve throughput and reduce delay and that it is possible to trade off delay for throughput and vice versa. This is in contrast to the general expectation that higher throughput can only be achieved with noticeable degradations in the end-to-end network delay. In this regard, we jointly formulate the end-to-end data rate allocation and adaptive channel coding (at the physical layer) within the general framework of network utility maximization (NUM). Depending on the choice of the objective function, we formulate two NUM problems: one aiming to maximize the aggregate network utility and another one aiming to maximize the minimum utility among the end-to-end flows to achieve fairness, which is of interest in certain vehicular network applications. Simulation results confirm that we can significantly decrease the average delay at the cost of a small decrease in throughput. This is achieved by maximizing the aggregate utility in the network when fairness is not the dominant concern. Furthermore, we also show that, even when resource allocation is performed to provide fairness, we can still decrease the maximum end-to-end delay of the network at the cost of a slight decrease in the minimum throughput.