ONSIDE-SELF: A Selfish Node Detection and Incentive Mechanism for Opportunistic Dissemination

The advent of IoT (the Internet of Things) has led to the necessity of fast and secure communication between devices, ranging from small sensors to top-of-the-line smartphones or laptops. One proposal for IoT communication is through 5G, which is estimated to be rolled out by 2020. However, the infrastructure for 5G communication might not always be present, or it should be avoided because of congestion. Moreover, employing it in smaller IoT networks can prove too expensive in some cases, while some small devices such as sensors might not even have 5G capabilities (or having them would greatly increase their price). For these reasons, opportunistic communication is an alternative for IoTs where mobile broadband connections cannot be used. Opportunistic networks are formed of mobile devices (such as smartphones and tablets belonging to social users) that communicate using close-range protocols such as Bluetooth or WiFi Direct. These networks are based on the store-carry-and-forward paradigm, where contacts between nodes are used opportunistically to transport data from a source to a destination, even though the two nodes might never be in direct communication range. Data dissemination assumes that nodes do not send directed messages (i.e., from a source to a pre-set destination), instead using channels to perform communication. Nodes are able to subscribe to channels, which are represented by interests (e.g., a node interested in “IT” will need to receive all messages marked with that tag). The main requirement of opportunistic networks is that the participating nodes should be altruistic, since communication is performed with the help of other nodes. However, this might not always be the case, since selfish nodes might decide that they do not want to help others. Such nodes should be detected and not allowed to participate in the dissemination process. This way, their messages will not be delivered, so they will be forced to become altruistic if they want a good networking experience. In this chapter, we propose a method for detecting and punishing selfish nodes in opportunistic networks dissemination, using gossiping mechanisms over the dynamic social network. Nodes learn about the behavior of other nodes and, when a contact occurs, share this information with an encountered device. We apply this method to an existing social and interest-based dissemination algorithm (ONSIDE) and show that it correctly detects and punishes selfish nodes, thus increasing the network’s behavior in terms of message delivery and congestion.