OVERLAY SPECTRUM REUSE IN A MULTICARRIER BROADCAST NETWORK: COVERAGE ANALYSIS Alberto Rico-Alvari˜ no, Carlos Mosquera, Fernando P´ erez-Gonz´ alez Signal Theory and Communications Department, Universidade de Vigo, 36310 Vigo, Spain email: {alberto,mosquera,fperez}@gts.uvigo.es ABSTRACT A secondary cognitive user overlaying its message in a broadcast multicarrier network is studied. The secondary user exploits the pri- mary message knowledge to convey its own information while pre- serving the primary user coverage area, determined by a bound on the BER, and taking into account the degradation due to the inser- tion of an echo in a dominant line of sight environment. The results are compared with those obtained when the coverage area is defined in capacity terms, which do not consider the practical degradation caused by the secondary replica of the primary message. 1. INTRODUCTION The overlay approach to a secondary user accessing a licensed band has arisen as an alternative to the interweave paradigm: while in the latter the cognitive user senses the spectrum, looking for unused fre- quency bands (the so-called white spaces) to use for the secondary transmission, the former tries to exploit the knowledge of the pri- mary message to transmit using the same frequency resources as the primary system while preserving the primary user Quality of Service (QoS) [1]. One of the scenarios where a secondary user could gain access to the primary signal is in broadcast Single Frequency Networks (SFN) [2], where the signal to be transmitted is delivered to the different primary transmitters via a distribution network. Therefore, the sec- ondary transmitter could join the primary network, acquire the pri- mary signal and exploit this knowledge to increase its spectral effi- ciency while preserving the primary user QoS. Thus, the insertion of a secondary signal is similar to the case of a transmitter inserting local content in a broadcast SFN using hier- archical constellations: the secondary (or local) transmitter allocates some of its available power to the primary (or global) message in order to meet an interference constraint, while the remaining power is used to convey the secondary (or local) information. Although this scenario has been previously studied in the litera- ture [3, 4] for the Additive White Gaussian Noise (AWGN) channel, the insertion of a secondary user conveying the primary message cre- ates a multipath channel with a potential performance degradation, especially in those systems with a strong Line Of Sight (LOS) com- ponent, with the AWGN channel the worst case scenario, as shown in [5]. Research supported by the European Regional Development Fund (ERDF) and the Spanish Government under projects DYNACS (TEC2010- 21245-C02-02/TCM) and COMONSENS (CONSOLIDER-INGENIO 2010 CSD2008-00010), and the Galician Regional Government under projects Consolidation of Research Units 2009/62 and 2010/85. In [2] it was shown that the prefiltering of the primary message at a cooperative secondary transmitter (a secondary transmitter that acts as a pure relay, without inserting its own information) dramati- cally improves the performance of the primary user. In the compan- ion paper [6] the insertion of a secondary message with a constraint on the Bit Error Rate (BER) of a single receiver is studied. In this paper we study the problem of inserting a secondary message while keeping the original coverage area of the primary transmitter, and compare the results with those obtained when the degradation due to the secondary replica is not taken into account. The paper is structured as follows: in Section 2 we present a capacity-based analysis of the primary coverage, while in Section 3 we propose an alternative analysis, based on a bound on the BER, and taking into account the possible degradation due to the inser- tion of a secondary replica in an AWGN channel. Finally, Section 4 presents the conclusions. 2. SIMPLIFIED SCENARIO In this section, we present a simple scenario where a secondary trans- mitter, who wants to maximize its own rate subject to a QoS con- straint regarding the primary service, is placed within the coverage area of a primary user. Let us assume that the primary user transmits at a fixed rate Rp with power Pp, and is located at xp ∈ R 2 . The capacity of the channel from the primary transmitter to a primary receiver lo- cated at xr is C (xr ) = log 2  1+ Ppl(xp,xr ) σ 2  bits per channel use, where l (x1, x2) denotes the propagation loss from a transmit- ter located at x1 to a receiver located at x2, and σ 2 is the noise power, assumed constant for all receivers. Thus, we define the initial coverage zone as C0 = {x | C (x) ≥ Rp}, or, equivalently, C0 =  x     Ppl(xp,x) σ 2 ≥ Υ0  where Υ0 =2 Rp − 1 is the required Signal to Noise Ratio (SNR) for a correct reception. We will assume that the secondary user has a fixed location in- side the coverage zone, xs ∈C0, and has a total transmit power Ps. This power has a twofold purpose, the reinforcement of the primary signal, for which γ 2 units of power are used, and the transmission of the secondary message, with a power consumption of ρ 2 , such that ρ 2 + γ 2 ≤ Ps. With the insertion of the secondary transmitter, the capacity of the channel of a primary receiver located at xr with the primary transmitter will be C ( xr ,γ 2 ,ρ 2 ) = log 2  1+ Ppl (xp, xr )+ γ 2 l (xs, xr ) ρ 2 l (xs, xr )+ σ 2  (1) where it has been assumed that the transmission of the primary signal