Let xi uv be the traffic load of flow i on link euv and xuv be the traffic load on link euv. Obviously, we have xuv = Q i=1 xi uv. Our objective is to route the traffic in the FiWi such that the network throughput is maximized.
Equations (2), (3) and (4) guarantee the flow conservation at each node in the network. Constraint (5) is the wireless link interference constraint in the network. Constraint (6) and (7) ensure that both xi uv and ëi are with non-negative value.
In order to measure the network throughput gain brought by the FiWi network, we compare the throughput obtained above to that in the traditional WMN. The throughput optimization problem for the traditional WMN can use the same LP formulation above on the graph G = (N,E) where N = {u0 }¾NG ¾NM and for any two gateways u, v ¸ NG, there are wireless links euv, evu ¸ E if and only if v is within ufs transmission range and vice versa as illustrated in Fig. 2.
Let xi uv be the traffic load of flow i on link euv and xuv be the traffic load on link euv. Obviously, we have xuv = Q i=1 xi uv. Our objective is to route the traffic in the FiWi such that the network throughput is maximized.Equations (2), (3) and (4) guarantee the flow conservation at each node in the network. Constraint (5) is the wireless link interference constraint in the network. Constraint (6) and (7) ensure that both xi uv and ëi are with non-negative value.In order to measure the network throughput gain brought by the FiWi network, we compare the throughput obtained above to that in the traditional WMN. The throughput optimization problem for the traditional WMN can use the same LP formulation above on the graph G = (N,E) where N = {u0 }¾NG ¾NM and for any two gateways u, v ¸ NG, there are wireless links euv, evu ¸ E if and only if v is within ufs transmission range and vice versa as illustrated in Fig. 2.
การแปล กรุณารอสักครู่..
