Device-centric communication in IoT: An energy efficiency perspective

Mushtaq Ahmad, Mudassar Ali, Muhammad Naeem, Ashfaq Ahmed, Muhammad Iqbal, Waleed Ejaz, Alagan Anpalagan

Research output: Contribution to journalArticlepeer-review

9 Scopus citations


Emerging device-centric communication technologies such as device-to-device (D2D) communication, devices-to-device (Ds2D) communication, and multi-homing (MH) D2D have been considered as essential parts of future fifth-generation networks as well as Internet of Things (IoT). The device-centric communication offers enhanced cellular data rates, high spectral efficiency, reduced latency, improved fairness, better energy efficiency, and extended coverage; however, the battery life of end devices is crucial to fully reap the benefits of this technology. In this article, a new method for device-centric communication in IoT system is introduced, where multiple source IoT devices (SIDs) can send data to multiple destination IoT devices (DIDs) using multiple radio resource blocks. This method is called devices-to-devices (Ds2Ds) communication. The objective is to select the optimal SIDs, DIDs, and radio resource blocks such that the total energy efficiency for all DIDs is maximized. A tree search algorithm is proposed to select the optimal SIDs, DIDs, and radio resource blocks. Extensive simulations have been carried out to compare energy efficiency per SID/DID for Ds2Ds, Ds2D, and MH-D2D. The simulation results show the superiority of Ds2Ds over Ds2D in terms of energy efficiency, which, in turn, implies better throughput. Furthermore, Ds2Ds is superior to MH-D2D in terms of energy consumption per source device, a very good and promising requirement for green communication.

Original languageBritish English
Article numbere3750
JournalTransactions on Emerging Telecommunications Technologies
Issue number2
StatePublished - 1 Feb 2020


Dive into the research topics of 'Device-centric communication in IoT: An energy efficiency perspective'. Together they form a unique fingerprint.

Cite this