Method of accelerating the shortest path problem
View Patent ↗The solution to the shortest path between a source node and multiple destination nodes is accelerated using a grouping of nodes, where the nodes are grouped based on distance from the source node, and a corresponding set of memory locations that indicate when a group includes one or more nodes. The memory locations can be quickly searched to determine the group that represents the shortest distance from the source node and that includes one or more nodes. Nodes may be grouped into additional groupings that do not correspond to the set of memory locations, when the distance from the source node to the nodes exceeds the range of memory locations. Advantageously, the disclosed system and method provide the ability to reach asymptotically optimal performance.
1. A computer implemented method of finding the shortest path between a source node and multiple destination nodes, said method comprising:
evaluating with a computer nodes tat neighbor a first node based on the distance from the source node to the neighboring nodes;
grouping with the computer the neighboring nodes into a plurality of groups based on the distance from the source node to the neighboring nodes;
initializing with the computer a set of memory locations, which contains one location for each group, by setting at least one bit in a memory location if the corresponding group contains at least one node;
selecting with the computer a group having at least one node, said group representing the shortest distance from the source node;
retrieving with the computer the next node from said selected group and deleting said node from said selected group.
2. The method of claim 1 , further comprising:
repeating the acts of evaluating, grouping, and initializing nodes using said next node as said first node.
3. The method of claim 1 , wherein grouping with the computer the neighboring nodes into a plurality of groups based on the distance from the source node to the neighboring nodes comprises:
storing a neighboring node in a first set of plurality of groups if the distance from said source node to said neighboring node is within a range;
storing a neighboring node in a second set of plurality of groups if the distance from said source node to said neighboring node is greater than said range;
wherein said set of memory locations contains one location for each group in said first set of plurality of groups.
4. The method of claim 1 , wherein said set of memory locations represents said range, said range being based on the number of locations in said set of memory locations.
5. The method of claim 4 , further comprising:
increasing the range represented by said set of memory locations; and
moving nodes from said second set of plurality of groups to said first set of plurality of groups when said nodes have an attribute that is within said increased range.
6. The method of claim 5 , wherein said increasing the range and moving nodes is performed when there are no nodes in said first set of plurality of groups.