Opportunistic Routing in Wireless Networks

Opportunistic routing is a technique, that makes use of the broadcasting nature of wireless networks i.e. when one node transmits, multiple other nodes can listen to the data transmitted. Rather than pre-selecting a single next hop, the route is determined dynamically based on which node is closest to the destination and the forwarding of data is done by the same. A core aspect of this method is determining the set of potential forwarders and ranking them based on certain parameters. This technique enhances the effectiveness, data transfer rate, and reliability of wireless networks.

Traditional routing consists of predetermined fixed paths, which might not always be optimal because of fluctuations in link quality, interference, and mobility and may not be available for transmission at all times. Unlike traditional routing, in Opportunistic Routing multiple paths are established and explored.

Elements of Opportunistic Routing

Candidates

These are the potential next-hop nodes. Various aspects, like closeness to the destination and link stability, influence candidate selection. Based on these metrics, the selection algorithm sets a priority of candidates and selects the next hop node.

The candidate set is a collection of nodes marked as potential forwarders.

Metrics

How metrics are used mainly depends on the network-related information present at each node. They can be classified into two categories:

1. Local Metric: The selection and prioritization of candidates depend on the local information of neighbors (e.g. the geographic position of nodes) or the link properties of neighbors (e.g. the link delivery ratio of the links).
2. End to End: In this case, the global information of the network topology is used to select and order the candidate Sets. The main factors involved are the characteristics of the wireless links and nodes. The calculated metric for each node is dependent on the cost of the remaining path from the neighbors of node to its destination.

Mathematics Behind Opportunistic Routing

Opportunistic Routing utilizes probabilistic and queuing theories. Network nodes become vertices, and links turn into edges in this representation. Connectivity information like link strength, distance, and signal strength are taken as the weights of the edges.

When packets arrive at a node, the edge weight metrics are used to probabilistically determine the next hop. Each node has a buffer queue where all the packets to be transmitted are queued. Based on the length of the queue and congestion, the next candidate metrics are altered.

Functioning

There are two main steps involved in how opportunistic routing works:

• Selection of potential forwarders between the source and destination
• Prioritization among these forwarders, where various metrics play a role.

As mentioned above, instead of pre-selecting a single specific node to be the next-hop as a forwarder for a packet, multiple nodes can potentially be selected. Thus the source can use multiple paths to deliver the packets to the destination.

When the current node transmits a packet, all the candidates that receive the packet successfully will coordinate with each other to determine which one would actually forward the packet, while the other nodes will simply discard the packet. This method minimizes the chance of multiple nodes transmitting the same packet, which could lead to collision.

Consider an example. The source node S has 3 intermediate nodes with packet delivery probability of 20%. Each node has packet delivery probability of 78% to the destination. Traditional routing will chose only one path for forwarding the data, while Opportunistic Routing will consider all the paths.

Illustration for Opportunistic Routing

Node S forwards data packets to multiple nodes using broadcasting. Each edge has a different probability weight based on metrics. Candidate nodes are {A, B, C}. Based on this set, each path from A to D is tried, whoever reaches the destination first is the chosen path. Notice, how the process takes place dynamically.

Forward Batch: It is a group of packets forwarded by a node to the next hop node.

Dynamic Relay Node can be used to forward the packet, to increase the transmission reliability and network throughput.

Popular Opportunistic Routing Protocols

• ExOR (Extremely Opportunistic Routing): ExOR broadcasts each packet, selecting a receiver to forward only after learning the set of nodes which really received the packet. It is among the first protocols to propose the idea of using a list of candidate forwarders and allowing the best receiver to take charge for forwarding.
• EEOR (Energy Efficient Opportunistic Routing): EEOR is an algorithm which works on the basis of selecting forwarders’ list and prioritizing the nodes in it. The expected cost of transmission is recorded against each forwarder node entry in the forwarders’ list.
• MORE (MAC-independent Opportunistic Routing and Encoding): It combines Opportunistic Routing with network coding to further increase throughput.
• EFFORT – EFFORT is an opportunistic routing protocol for wireless sensor networks (WSNs), based on the Opportunistic End-to-end Cost (OEC) metric, which evaluates the energy cost of data transmissions. It comprises three main components: OEC computation, selection of candidates with relay priorities, and data forwarding with OEC updating.

Advantages

• Fault tolerance – With multiple forwarding options, there’s a better chance of delivery, even with unstable links.
• Adaptability – Ability to handle dynamic changes and fluctuations in the network.
• Improved Throughput – By leveraging the broadcast nature of the wireless medium and allowing multiple candidates to forward packets, Opportunistic Routing can achieve higher throughput than traditional protocols.

Limitations

• Implementation Complexity – The dynamic nature and the need for coordination make Opportunistic Routing protocols more complex. Also requires additional security mechanism.
• High network overhead – The need for coordination among candidate forwards and maintaining candidate lists can cause overhead.
• Higher Energy Consumption – Consumes more energy as packets are forwards to multiple nodes.
• There is a chance of packet duplication.

Frequently Asked Questions

Q.1: What is Opportunistic Routing and how does it differ from traditional routing methods?

Answers:

Opportunistic Routing is a routing technique that capitalizes on the broadcasting nature of wireless networks, allowing multiple potential next hop nodes to listen in when a node transmits data. The actual route is dynamically determined by the node that is closest to the destination. Unlike traditional routing, which uses predetermined fixed paths, Opportunistic Routing explores multiple paths to find the most optimal one, adjusting to changes in link quality, interference, and mobility.

Q.2: How are candidates selected in Opportunistic Routing and what metrics influence their prioritization?

Answer:

Candidates are potential next hop nodes. Their selection is influenced by factors like their proximity to the destination and link stability. Metrics used in this process can be classified into two categories:

• Local Metric: This is based on local information of neighbors, such as their geographic positions or link properties like link delivery ratios.
• End to End: This uses global information about the network topology. Factors include the characteristics of wireless links and nodes, and the metric for each node depends on the cost of the remaining path from a node’s neighbors to its destination.

Q.3: What are some of the popular Opportunistic Routing Protocols and their features?

Answer:

Some popular Opportunistic Routing Protocols include:

• ExOR (Extremely Opportunistic Routing): One of the pioneering protocols, ExOR introduces the concept of using a list of candidate forwarders and allowing the best receiver to take the initiative for forwarding.
• MORE (MAC-independent Opportunistic Routing and Encoding): This protocol combines Opportunistic Routing with network coding to further boost throughput.