1 Joint Precoding and On-Board Beamforming for Multiple gateway Multibeam Satellite Systems Vahid Joroughi, Sina Maleki, Bhavani Shankar M.R., Symeon Chatzinotas, Joel Grotz and Bj¨ orn Ottersten Abstract— This paper aims to design joint precoding and on- board beamforming of a multiple gateway multibeam satellite system, either in a hybrid space-ground mode, or in a totally on-board one. In such an architecture, with employing high throughput full frequency reuse pattern over both user and feeder links, each gateway serves a cluster of adjacent beams such that the adjacent clusters are served through a set of gateways that are located at different geographical areas. However, such a system brings in two challenges to overcome. First, the interference in both user and feeder links is the bottleneck of the whole system and applying interference mitigation techniques becomes necessary. Second, as the data demand increases, the ground and space segments should employ extensive bandwidth resources in the feeder link accordingly. This entails embedding an extra number of gateways aiming to support a fair balance between the increasing demand and the corresponding required feeder link resources. To solve these problems, this study investigates the impact of employing a joint multiple gateway architecture and on-board beamforming scheme. It is shown that by properly designing the on-board beamforming scheme, the number of gateways can be kept affordable even if the data demand increases. Moreover, Zero Forcing (ZF) precoding techniques are considered to cope with the interference in both user and feeder links which embed in the following premises: (i) each gateway constructs a part of block ZF precoding matrix, (ii) the satellite and gateways perform the precoding scheme, and (iii) a joint design of ZF precoding and on-board beamforming at the payload of the satellite so that no signal processing scheme is conceived at the gateways. The provided simulation results depict the performance gain obtained by our proposed schemes. Index Terms—Multibeam satellite systems, multipe gateway systems, on-board beamforming, precoding techniques. I. I NTRODUCTION A. Motivation Fixed broadband satellite systems are an integral part of the communications technology, aiming to provide ubiquitous broadband services. Built on the Multiuser Multiple-Input Multiple-Output (MU-MIMO) framework, the use of multiple spot beams in modern broadband satellites have been recently considered by employing fractional frequency reuse among beams [1]. Such systems rely on employing a large number of spot beams instead of a single (global) beam in the coverage area to provide higher spectral efficiency [2]. However, one of the major challenges in multibeam architec- ture is how to deal with interference in the access network. Indeed, adjacent beams create high levels of interference due to the side lobes of the radiation pattern of beams on the Earth surface. Therefore, typically adjacent beams operate on different frequency bands. In this context, N c is the essential parameter which corresponds to the number of disjoint fre- quency bands employed on the coverage area (N c ≥ 1). Another promising technique is to use full frequency reuse pat- tern (N c =1) by resorting interference mitigation techniques. In this way, interference mitigation techniques as precoding in the forward link and multi-user detection in the return link have been proposed in the past [3]-[4]. Apart from the already mentioned interference limitation, another major challenge of multibeam systems is to deal with the large spectral demands of the Feeder Link (FL), i.e. the bidirectional link between satellite and the Gateway (GW), whose bandwidth requirements increase as it aggregates the traffic of all users. Keeping a full frequency reuse allocation (N c =1), the required FL resources can be calculated as B feeder-link = N B beam , (1) where N is the number of on-board feed signals. The notations B beam and B feeder-link are the per-beam and the FL required bandwidths, respectively. Let us consider a total number of K beams with N>K. From (1), it is evident that any user/beam available bandwidth enhancement forces the FL resources to be increased accord- ingly and, eventually the FL might become the communication bottleneck. Note that, in contrast to the single feed per beam architectures, i.e. N = K, applying Multiple Feeds per Beam (MFB) at the payload, i.e. N>K, can reduce the scan losses for a large continental coverage areas, and are specially suited for contour beams [1]. Recently, some techniques have been proposed in order to reduce the FL spectrum requirements. One solution is moving the FL from the Ka band to the Q/V band so that a larger available bandwidths can be achieved [5]. Unfortunately, the Q/V carrier frequencies suffer the impact of an extremely large fading and more advanced transmitting schemes at the GW are needed. Another solution is to employ a Beamforming Network (BFN) at the payload aiming at: i) synthesizing the amplitude and phase modulating the excitation of each on-borad feed in the MFB scheme [6], ii) reducing the FL bandwidth requirements [7] by B feeder-link-onboard = KB beam , (2) where B feeder-link-onboard denotes the FL resources that is required after employing the on-board BFN with B feeder-link-onboard < B feeder-link and N > K. However, the FL resources B feeder-link-onboard must be increased with the number of beams, i.e. K. Besides, the volume and calibration requirements of any on-board BFN are currently its main drawback. Another promising option is the use of on-ground multiple GW architecture. This architecture exploits the multiplexing diversity by reusing all the available FL bandwidth across the different Gateways (GWs) [8]-[9]. In this context, the required FL bandwidth becomes B feeder-link-MG = N F B beam , (3) where F is the number of GWs, and B feeder-link-MG denotes the FL resources which is required at multiple GW architecture. Indeed, the multiple GW architecture reduces the required FL