Research algorithm can speed IoT traffic

  • February 7, 2024
  • Steve Rogerson
Possible multipath routing structure.

Mathematicians at RUDN University in Moscow have created a routing algorithm that can speed IoT traffic.

The algorithm optimally splits traffic, which improves network speed and reliability.

The architecture of IoT networks faces multiple requirements such as scalability, flexibility, reliability and availability. To provide all these qualities, efficient routing is necessary.

The mathematicians proposed using a scheme in which transmission occurs in several ways at once; the traffic flow is split and transmitted through various intermediate nodes. The multipath routing method itself is not new, but the question is the optimal choice of specific traffic paths.

There are several approaches to choosing routes for different networks. In some cases, for example, a random selection of paths is used. However, for the IoT network, classical methods are not effective.

“Multipath routing allows data to be transferred faster than single-path routing,” said Ammar Muthanna, head of the RUDN University wireless network simulation centre. “The number of possible routes in IoT networks can be large. To choose routes, one needs to solve the problem of traffic distribution.”

The proposed method for multipath routing is based on dynamic programming, in which a problem is broken down into simpler subtasks. This reduces the computation time. The created algorithm was compared with random uniform traffic distribution and other classical approaches.

The method outperforms classical path selection methods in multipath routing, especially for dense IoT networks. The number of intermediate nodes included in the route was reduced by 34%. Computational costs were reduced by 52% and data delivery time by 40%.

“We created a method for selecting optimal routes for the best data transfer that uses dynamic programming,” said Muthanna. “The proposed method increased the efficiency of using network resources. The route selection scheme increased the transmission rate by 40% compared with random traffic distribution. In addition, our algorithm has surpassed the classical ones in terms of energy consumption, packet delivery delay, packet delivery ratio and costs.”

The paper was published in the journal Mathematics (www.mdpi.com/2227-7390/11/19/4168).