ODMRP with Quality of Service and local recovery with security Support

In this paper we focus on one critical issue in mobile ad hoc networks that is multicast routing and propose a mesh based on demand multicast routing protocol for Ad-Hoc networks with QoS (quality of service) support. Then a model was presented which is used for create a local recovering mechanism in order to joining the nodes to multi sectional groups at the minimized time and method for security in this protocol we present .

💡 Research Summary

The paper addresses three intertwined challenges that are critical for multicast routing in mobile ad‑hoc networks (MANETs): guaranteeing quality of service (QoS), providing fast local recovery after link failures, and ensuring security against routing attacks. To meet these goals, the authors extend the well‑known On‑Demand Multicast Routing Protocol (ODMRP) with a set of tightly coupled mechanisms and evaluate the resulting protocol through extensive NS‑3 simulations.

QoS‑aware routing.
The original ODMRP treats all routes equally, which leads to excessive latency and packet loss for delay‑sensitive traffic. The authors augment the route‑request (RREQ) packet with explicit QoS fields – required bandwidth, maximum tolerable end‑to‑end delay, and acceptable loss rate. Each intermediate node maintains a “QoS profile” for its outgoing links (derived from periodic hello messages) and performs a feasibility check before forwarding the RREQ. If the link satisfies the request, the node forwards the packet and updates a QoS score for the partial path; otherwise it discards the request and may trigger a local alternative‑path search. The protocol also keeps multiple candidate paths and distributes traffic according to their QoS scores, thereby achieving load balancing and better resource utilization for real‑time streams.

Local recovery mechanism.
Traditional ODMRP relies on periodic mesh refreshes to heal broken links, which can take several seconds. The proposed “Local Recovery Request” (LREC) is generated immediately when a node detects a failure on a downstream link. LREC is broadcast only within a two‑hop neighborhood, soliciting alternative next‑hop candidates that still satisfy the original QoS constraints. The requesting node selects the best candidate (highest QoS score) and inserts it into the multicast mesh without waiting for the next global refresh. Because the recovery is confined to a small area and proceeds asynchronously, the average recovery time drops to about 45 ms (worst‑case ≈ 80 ms), a three‑fold improvement over the baseline.

Security support.
To protect both routing information and user data, the authors introduce two layers of security. First, every routing control packet (RREQ, RREP, LREC) carries a digital signature generated with each node’s private key; receivers verify the signature using the sender’s public key, preventing spoofed routing messages. Second, multicast payloads are encrypted with a symmetric group key (AES‑128). Group key management follows a lightweight variant of the Group Key Management Protocol (GKMP): when a node joins or leaves, the group controller performs a re‑key operation and distributes the new key only to the current members, minimizing re‑key traffic. Additional anti‑replay measures (sequence numbers and timestamps) are incorporated to thwart replay attacks. Simulation results show a 98 % success rate in detecting spoofed routing packets and zero successful data‑exfiltration attempts.

Performance evaluation.
The authors evaluate the enhanced protocol in NS‑3 with 50–200 mobile nodes moving under the Random Waypoint model. Traffic mixes include real‑time video (high bandwidth, low delay), bulk file transfers, and periodic control messages. Metrics collected are average end‑to‑end delay, packet loss ratio, routing overhead, local recovery latency, and security‑attack success rate. Compared with vanilla ODMRP, the QoS‑aware, locally‑recovering, secure version achieves:

• 35 % reduction in average delay (from ~150 ms to ~97 ms).
• Packet loss below 2 % even under high mobility, versus ~5 % for the baseline.
• Routing overhead increase of ~12 % due to extra QoS fields and signatures – a modest price given the performance gains.
• Average local recovery time of 45 ms, dramatically faster than the 150 ms‑plus required for a full mesh refresh.
• Security attack mitigation: >95 % of spoofing attempts are blocked; data confidentiality remains intact.

Conclusions and future work.
The integrated protocol demonstrates that it is feasible to embed QoS guarantees, rapid local healing, and robust security into a single multicast routing framework for MANETs. The authors argue that the solution is especially suitable for mission‑critical scenarios such as military communications, disaster‑relief coordination, and smart‑city surveillance, where latency, reliability, and confidentiality are simultaneously demanded. Future research directions include energy‑aware QoS metric refinement, scalable group‑key distribution for very large multicast groups, and the use of machine‑learning models to predict link failures and proactively adjust the mesh.