Equalized network latency for multi-player gaming
View Patent ↗A method is provided for equalizing latency among a group of connections that are being used by the same latency-sensitive application, such as for multi-player network gaming. Delay is added to the packet flow in each connection as necessary at the provider edge router connecting the application server to the network. An NMS connected to the provider edge router continuously monitors the round-trip delay of each connection, determines the additional delay of each connection required to equalize the delays, and informs the provider edge router. The provider edge router buffers traffic on each connection for the respective duration as indicated by the NMS. The NMS may also reroute some of the connections over higher latency resources to reduce the amount of buffering required, which provides the additional advantage of using what might otherwise be less used or economically less valued resources.
1. A method of using a network management system (NMS) to reduce a latency disparity between a plurality of connections accessing a server through a provider edge device, wherein the NMS manages the provider edge device, the method comprising:
determining an inherent latency of the plurality of connections with the NMS;
designating a largest inherent latency of the determined inherent latencies as a target latency, wherein dynamic adjustment of the target latency occurs;
for the plurality of connections, attempting to reroute a connection when an original connection has an initial inherent latency lower than the target latency by an amount equal to or greater than a threshold;
for each rerouting attempt, selecting a new route having a new latency greater than the initial inherent latency of the original connection;
for each new route, reestablishing the original connection along the new route, thereby creating a new connection; and
determining a new inherent latency of each new connection.
2. The method of claim 1 , further comprising:
determining an imposed latency for each new connection, each imposed latency being a difference between the new inherent latency of the new connection and the target latency; and
buffering packets for each new connection for a duration of time equal to the imposed latency of the new connection.
3. The method of claim 2 , wherein determining the new inherent latency and determining the imposed latency are repeated at intervals, whereby the duration for which packets are buffered for a new connection may vary with time.
4. The method of claim 3 , wherein determining the new inherent latency of each new connection further comprises:
determining a round trip delay between an end user device of the new connection and the provider edge device through which the server accesses the network.
5. The method of claim 4 , wherein determining the round trip delay further comprises:
the provider edge device pinging the end user device of the new connection.
6. The method of claim 2 , further comprising:
using rerouting to perform coarse equalization of latencies; and
using buffering to perform fine equalization of latencies.
7. The method of claim 2 , further comprising:
informing the provider edge device of the imposed latency for each new connection with the NMS.
8. The method of claim 7 , further comprising:
maintaining, in the provider edge device, values of the imposed latencies.
9. The method of claim 7 , further comprising:
placing, with the provider edge device, packets traveling from each connection to the server, in a buffer having a delay equal to the imposed latency.
10. The method of claim 7 , further comprising:
placing, with the provider edge device, packets traveling from the server to each connection, in a buffer having a delay equal to the imposed latency.
11. The method of claim 7 , further comprising:
placing, with the provider edge device, all packets traveling between the server and each connection, in a buffer having a delay equal to half of the imposed latency.
12. The method of claim 1 , wherein determining the new inherent latency of each new connection further comprises:
determining a round trip delay between an end user device of the new connection and the provider edge device through which the server accesses the network.
13. The method of claim 12 , wherein determining the round trip delay further comprises:
the provider edge device pinging the end user device of the new connection.
14. The method of claim 1 , further comprising:
using rerouting to perform coarse equalization of latencies; and
using buffering to perform fine equalization of latencies.