Propagation Measurements for D2D in Rural Areas Sathyanarayanan Chandrasekharan † , Akram Al-Hourani † , Kagiso Magowe † , Laurent Reynaud ⋆ and Sithamparanathan Kandeepan † † School of Electrical and Computer Engineering, RMIT University, Melbourne, Australia ⋆ Orange, Lannion, France Email: s.chandrasekharan.2014@ieee.org Abstract—Device-to-Device (D2D) communications is consid- ered to be a promising technique in next-generation communica- tion networks aiming to enhance spectrum utilization and energy- efficiency. In this paper, we investigate the radio propagation behavior of the D2D channel in rural areas by conducting field measurements. We characterize the D2D channel by estimating the path-loss exponent and the standard deviation of the shad- owing and by presenting the shadowing correlation analysis in various propagation environments using ISM band frequencies of 922 MHz and 2,466 MHz. The reported results are of great importance to system designers and researchers alike, to assist the development and the evaluation of innovative D2D protocols under realistic propagation conditions. Keywords—Device-to-Device communication, radio propaga- tion, rural areas, path-loss, regional area networks. I. I NTRODUCTION Device-to-Device communication (D2D) is the ability of peers to interchange data directly without the need for a cellular base station to handle and route the traffic. This technology has recently gained a special focus with the con- tinuing development of LTE-Advanced and cellular networks towards 5G networks [1]. The demand to accommodate D2D technology can be mainly ascribed to public safety agencies [2], [3] as well as to commercial cellular operators, since the anticipated gains will significantly enhance the network availability and the spectral efficiency. When D2D devices are in close proximity, the required transmit power is much lower than normally needed to com- municate with a cellular base station, thus leading to lower in- terference in the cellular network, and a higher energy saving. Moreover, a D2D-enabled device is expected to have cognitive radio features [4] and can act as a mobile relay assisting other devices to access the cellular network, a scenario that can reduce network deployment costs, as well as, provide an essential backup in case of major network failure due to a natural disaster [5]. Fig. 1 depicts the four major scenarios for using D2D technology identified by 3GPP [6], the first scenario I-A shows a D2D pair out of the reach of the main cellular network, a situation that is frequently faced by public safety personnel (for example during a fire fighting operation in remote bush fire). In scenario I-B one device acts as a mobile relay aiming to extend cell-edge coverage. While scenarios I- C and I-D represent the case when D2D communication takes place underlaying a cellular coverage where a local source- sink type traffic is occurring between the two nearby devices (such as local voice/video calls, file transfer, etc...). When applying any of the communication scenarios, a radio channel model is required to study the various radio dynamics in the network [7]. Namely, to model the expected path-loss between a communicating D2D pair and how much interference it might cause to other D2D devices. In this paper we shed light on D2D path-loss behaviour in rural areas, generally characterized by sparse man-made structures and low user density. Specifically, we report the results of a measurement campaign conducted over ISM band frequencies of 922 MHz and 2, 466 MHz in a typical Australian country side (rural area), covering three environments with two test areas in each environment. The first environment consists of man-made structures on both sides of the road, the second environment consists of man-made structures on one side of the road and vegetation to the other side, and the third environment consists of a forest with tall tress and bushes. The first two en- vironments are inhabited areas with low user density whereas the last environment is unoccupied by humans but is quite important in case of a bush fire scenario where fire fighters are expected to heavily depend on wireless communication to coordinate their operations. Such emergency scenarios are being investigated with respect to D2D in the EU funded FP7 project ABSOLUTE [8] [9]. We present the measurements in terms of the observed statistical parameters of the collected samples, where we follow the well-known log-normal shadowing model [10]. The reported parameters are namely, the path-loss exponent and the standard deviation of the shadowing component. Also, we present the shadowing correlation analysis. These results will be useful for researchers to develop and evaluate D2D protocols and algorithms based on D2D such as [11], [12], [13], [14], [15], [16] and [17] under realistic propagation conditions. The rest of this paper is structured as the following, in section II we present the recent work related to D2D radio channel characterization. In section III we describe the setup of the measurement and the utilized tools, also we describe the environments and the test areas where the measurement was conducted. In section IV we report the statistical results of the conducted measurement campaign. And in section V we provide the concluding remarks and future research paradigms. II. RELATED WORKS Prior research on empirical measurements based character- ization of propagation channels ranging from cellular, D2D, V2V (Vehicle-to-Vehicle) and millimetre-wave communica- tions, can be found in literature [18], [19], [20], [21], [22]. The measurement campaigns were conducted in different propaga- tion conditions to characterize different propagation phenom- ena (path-loss, small/large-scale fading, time-delay spread),