This paper presents SimBet Routing, a novel routing protocol for disconnected delay-tolerant Mobile Ad Hoc Networks (MANETs) that leverages social network analysis techniques. The key challenge in sparse MANETs is to find a route that provides good delivery performance and low end-to-end delay in a disconnected network where nodes may move freely. Traditional routing protocols like AODV, DSR, DSDV, and LAR assume a fully connected network and fail in sparse MANETs. SimBet Routing addresses this by using social network analysis to identify bridge nodes based on their centrality and social similarity to the destination node. SimBet Routing exploits the exchange of pre-estimated 'betweenness' centrality metrics and locally determined social 'similarity' to the destination node. It uses ego networks to determine a node's centrality without requiring complete knowledge of the entire network. The algorithm makes no assumptions of global knowledge and forwarding decisions are based solely on local calculations. SimBet Routing is evaluated using real trace data from the MIT Reality Mining project and compared to Epidemic Routing and PRoPHET Routing. SimBet Routing achieves delivery performance close to Epidemic Routing but with significantly reduced overhead. It outperforms PRoPHET Routing, particularly when the sending and receiving nodes have low connectivity. The performance of SimBet Routing is evaluated in terms of total number of messages delivered, average end-to-end delay, average number of hops per message, and total number of forwards. SimBet Routing performs well in all metrics, especially in scenarios where nodes have low connectivity. The results show that SimBet Routing is effective in finding routes in sparse MANETs without the additional overhead of redundantly forwarding messages. The paper concludes that SimBet Routing is a promising approach for routing in disconnected delay-tolerant MANETs.This paper presents SimBet Routing, a novel routing protocol for disconnected delay-tolerant Mobile Ad Hoc Networks (MANETs) that leverages social network analysis techniques. The key challenge in sparse MANETs is to find a route that provides good delivery performance and low end-to-end delay in a disconnected network where nodes may move freely. Traditional routing protocols like AODV, DSR, DSDV, and LAR assume a fully connected network and fail in sparse MANETs. SimBet Routing addresses this by using social network analysis to identify bridge nodes based on their centrality and social similarity to the destination node. SimBet Routing exploits the exchange of pre-estimated 'betweenness' centrality metrics and locally determined social 'similarity' to the destination node. It uses ego networks to determine a node's centrality without requiring complete knowledge of the entire network. The algorithm makes no assumptions of global knowledge and forwarding decisions are based solely on local calculations. SimBet Routing is evaluated using real trace data from the MIT Reality Mining project and compared to Epidemic Routing and PRoPHET Routing. SimBet Routing achieves delivery performance close to Epidemic Routing but with significantly reduced overhead. It outperforms PRoPHET Routing, particularly when the sending and receiving nodes have low connectivity. The performance of SimBet Routing is evaluated in terms of total number of messages delivered, average end-to-end delay, average number of hops per message, and total number of forwards. SimBet Routing performs well in all metrics, especially in scenarios where nodes have low connectivity. The results show that SimBet Routing is effective in finding routes in sparse MANETs without the additional overhead of redundantly forwarding messages. The paper concludes that SimBet Routing is a promising approach for routing in disconnected delay-tolerant MANETs.