Topology aware MANET for mobile networks
View Patent ↗Systems and methods provide adaptability in a mobile ad hoc network (MANET), based on an existing protocol, such as adaptive hybrid domain routing (AHDR). The systems and methods are especially suited for fast changing topologies that may change after a reactive route discovery has been completed.
1. A method for adaptively selecting a domain lead node from among a set of nodes in a mobile ad hoc network, the method comprising:
(a) establishing a link change threshold percentage;
(b) establishing a threshold delta link state level (LSL) increment value;
(c) calculating, by each node in the network, for each node in the network what percentage of links have changed by more than the threshold delta link state level (LSL) increment value;
(d) determining, by each node in the network, if the calculated percentage of changed links for the given node exceeds the link change percentage threshold;
(e) adaptively computing, by each node in the network, a node coverage value (NC) for each node from among the set of nodes in the network by adaptively adjusting a state value, s 0 , for each node, downward, in response to satisfying said determining step (d); and
(f) selecting, by the set of nodes in the network, said domain lead node from among said set of nodes in said network based on a highest computed node coverage value (NC).
2. A method according to claim 1 , wherein said determining step (d) is performed at successive update increments.
3. A method according to claim 1 , wherein the link change threshold percentage is adjusted in accordance with a type of network.
4. A method for adaptively selecting a domain lead node from among a set of nodes in a mobile ad hoc network, the method comprising: for a given node from among said set of nodes, computing, by the given node, a node coverage (NC) value by:
(a) adaptively adjusting a plurality of weight values for neighbor nodes seen by said given node, wherein said adaptive adjustment is based upon at least one of a current state of said neighbor nodes and a current link state level (LSL) of said neighbor nodes; and
(b) selecting, b set of nodes in the network, said domain lead node from among said set of nodes based on a given node having a highest computed node coverage value (NC),
wherein said node coverage value for node A is adaptively computed as:
NC( A )= s 0 *[w 0 *n +(w 1 *s 1 +w 2 *s 2 +w 3 *s 3 + . . . +w n *s n )]
where
s 0 =a state of node A,
s 1 . . . s n =a state of each neighbor node,
n=a total number of nodes within 1 hop around node A,
w 0 =an adaptive weight put on a number of nodes, and
w 1 . . . w n =weights of each neighbor node depending on the state and the LSL.
5. A method according to claim 4 , wherein said weight values w n are computed as:
w n =(LSL n *min(1,max(0,LSL n −a )))+( z *max(0,(LSL n −(15− b ))))
where
a=sets how many of the bottom LSLs get zeroed,
b=sets how many of the top LSLs get scaled up,
n is a positive integer, and
z=weight of higher LSL.
6. A method for adaptively selecting a domain lead node from among a set of nodes in a mobile ad hoc network, the method comprising: for a given node from among said set of nodes, computing, by the given node, a node coverage (NC) value by:
(a) adaptively adjusting an adaptive weight w 0 parameter in accordance with the number of neighbor nodes N seen by said given node; and
(b) selecting, by the set of nodes in the network, said domain lead node from among said set of nodes based on a given node having a highest computed node coverage value (NC),
wherein said node coverage value for a node A is adaptively computed as:
NC( A )= s 0 *[w 0 *n +(w 1 *s 1 +w 2 *s 2 +w 3 *s 3 + . . . +w n *s n )]
where
s 0 =a state of node A,
s 1 . . . s n =a state of each neighbor node,
n=a total number of nodes within 1 hop around node A,
w 0 =an adaptive weight put on a number of nodes, and
w 1 . . . w n =weights of each neighbor node depending on the state and the LSL.
7. A method for adaptively selecting a domain lead node from among a set of nodes in a mobile ad hoc network, the method comprising: for a given node from among said set of nodes, computing, by the given node, a node coverage (NC) value by:
(a) adaptively adjusting a plurality of weight values for neighbor nodes seen by said given node, wherein said adaptive adjustment is based upon at least one of a current state of said neighbor nodes and a current link state level (LSL) of said neighbor nodes; and
(b) selecting, by the set of nodes in the network, said domain lead node from among said set of nodes based on a given node having a highest computed node coverage value (NC),
wherein said weight values w n are computed as:
w n =(LSL n *min(1,max(0,LSL n −a )))+( z *max(0,(LSL n −(15 −b ))))
where
a=sets how many of the bottom LSLs get zeroed,
b=sets how many of the top LSLs get scaled up,
n is a positive integer, and
z=weight of higher LSL.