In-Network Processing on Low-Cost IoT Nodes for Maritime Surveillance

Commercially available system of distributed wireless sensors could increase the Navy's intelligence collection footprint.

The effective distribution of offensive weapon capabilities to naval units at the tactical edge is a critical focus for Navy leaders. A direct byproduct of this priority is the need to employ sensor and data collection systems that can effectively guide the targeting of that offensive capability. In the recent past, wireless sensor networks have received limited use in the maritime domain due to the exploratory nature of technology, high system complexity and the high cost of system deployment. With the Internet-of-Things revolution, commercially available hardware and software components can be used to build low-cost, reliable, disposable wireless sensor networks that can leverage in-network processing schemes to greatly expand the intelligence collection footprint.

Centralized Sensor Network

The demonstrated application of the tactical low-cost sensor network (TLCSN) is the collection, processing, and dissemination of 802.11 messages—e.g. beacons, probe requests and data packets. This data provides enhanced situational awareness, identifying and displaying what clients and APs may exist in a specific operating area, as well as add context by analyzing and characterizing any trends and patterns of movement that may appear.

The overall system assists in characterizing the wireless environment by aggregating the data into a streamlined format for display. In colloquial terms, the data set is comparable to that provided by “war-driving”; however, this data serves as an initial demonstration of the sensor node and network capabilities. It provides enough volume over time to serve as an interesting data set for analysis while also providing useful data to a commander. Various other sensors can be added for different intelligence gain and used either interchangeably or in concert with each other.

There are four main components of the TLCSN system: the sensor nodes, the server/broker platform, the system controller, and the overall network structure.

Sensor Nodes

The sensor nodes conduct the actual data collection and reporting. They would be spread across a specified target area as dictated by the operational or testing requirements. When deployed, they would be automatically activated and establish connectivity with the network at large. Through this network connection, each node will be able to transmit and receive collected data across the network, reach back to a designated repository, and receive additional or modified tasking by a system controller. Nodes are assumed to be more or less static during their collection mode in this environment. Minimal movement, accounting for drift and currents, is acceptable.

Server / Broker Platform

The broker consolidates, processes and distributes data across the network as required. It has a complex role and provides various capabilities from acting as a VPN server allowing connectivity from outside the network, to serving as the broker for collected data. In some situations, it may also act as a WAP and DHCP server for the local network.

System Controller

The system controller serves as the user interface for the system. It provides mechanisms for accessing, viewing and analyzing the data, while providing additional tasking and control to various sensors. For example, if the controller wishes to use the sensor node for additional activities beyond the passive collection of data, that capability is available. To provide the maximum flexibility for this system, it is assumed that the controller resides at some external location. In all cases, it connects to the sensor network via a VPN connection and is never physically present on the same network.

Network

There are three general network architectural constructs that are explored. For the sake of simplicity, each one is illustrated using only five nodes. Actual implementation can be scaled to a much larger extent—both in the number of nodes working with a server (covering more area within a general region), as well as the total number of servers (covering more regions). This would serve to provide both greater breadth as well as depth of coverage.

This work was done by Andrew R. Belding for the Naval Postgraduate School. NPS-0006



This Brief includes a Technical Support Package (TSP).
Document cover
In-Network Processing on Low-Cost IoT Nodes for Maritime Surveillance

(reference NPS-0006) is currently available for download from the TSP library.

Don't have an account? Sign up here.