MANET Routing with Independent Dominating Set Clusterheads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing MANET routing protocols like OLSR face challenges with excessive routing overhead and bandwidth requirements due to the identification of multi-point relays (MPRs), which becomes costly as network density increases, and the NP-complete set-cover problem of selecting MPRs without global coordination.
Innovation Solution
Implementing an independent dominating set (IDS) based routing by designating a subset of clusterhead nodes within the network, using passive clustering procedures to transition node statuses, and transmitting adaptive hello messages without neighbor lists, followed by link status advertisements and distance vector distributions to minimize routing overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If multi-point relays (MPRs) are selected to restrict duplicate retransmissions in OLSR, then routing control overhead is reduced, but the number of selected MPRs may be as high as the number of nodes in the MANET, making link discovery increasingly expensive as network density increases
Solution Approach 1:
The patent extracts the MPR selection problem from the set-cover formulation and replaces it with an independent dominating set (IDS) approach. By removing nodes from the MPR set until no two MPRs are adjacent (independent set constraint), the patent reduces the number of nodes required while maintaining coverage of all non-MPR nodes. This extraction of redundant MPRs directly addresses the contradiction by reducing both the quantity of MPR nodes and the associated routing control overhead.
Solution Approach 2:
Instead of starting with all nodes as MPRs and trying to minimize the set (the conventional set-cover approach), the patent inverts the approach by starting with a dominating set and removing nodes to create an independent set. This inversion transforms the NP-complete set-cover problem into a more tractable IDS problem that naturally reduces the number of relay nodes while maintaining network coverage.
2Reliability
If hello messages include a list of all one-hop neighbors, then nodes can discover two-hop neighbors effectively, but routing overhead and bandwidth requirements increase significantly as network density increases
Solution Approach 1:
The patent applies local quality by making hello messages from MPR nodes different from ordinary node hello messages. MPR nodes include their one-hop neighbor lists in hello messages, while non-MPR nodes omit this information. This localized differentiation allows non-MPR nodes to discover two-hop neighbors through MPR nodes without every node needing to broadcast complete neighbor lists, thereby reducing overall routing overhead while maintaining discovery accuracy.
Solution Approach 2:
MPR nodes serve as intermediaries that collect and advertise link status information from their one-hop neighbors. Instead of every node broadcasting its neighbor list (which would create excessive overhead), MPR nodes act as mediators that aggregate this information and propagate it network-wide through their own hello messages and link status advertisements, efficiently enabling two-hop neighbor discovery without proportional increase in overhead.
3Reliability
If the number of selected MPRs increases to cover all nodes, then link status coverage is improved, but overlapping link advertisements become significant and hello overhead increases
Solution Approach 1:
The patent extracts redundant MPR nodes from the selection by enforcing the independent set constraint. By removing MPR candidates that are adjacent to other MPRs, the patent eliminates overlapping coverage areas where multiple MPRs would advertise the same link status information. This extraction reduces the number of nodes performing expensive hello message exchanges and link status advertisements, directly reducing energy consumption and overhead while maintaining sufficient link status coverage through the remaining independent MPRs.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
A communications node (102) of a multi-node communications network (100) is disclosed. In embodiments, the communications node includes a communications interface (108) and controller (104). The controller transitions the clustering status of the communications node to a clusterhead node status. The clusterhead node identifies a cluster of neighboring nodes with which it is in communication by transmitting hello messages identifying the clusterhead node and its status but omitting a one-hop neighbor list. The clusterhead node refines link discovery to the nodes of the multi-node communications network by flooding the network (e.g., via its cluster of neighboring nodes) with routing status messages, which network nodes not part of the dominating set of clusterhead nodes are restricted from sending.