The need for Tower Mounted Amplifiers

Transcription

1 The need for Tower Mounted Amplifiers João Moreira Rebelo and Nuno Borges Carvalho and Instituto de Telecomunicações, Universidade de Aveiro, Portugal Introduction Network planning in GSM systems is moving towards re-engineering, that means the actual problems, in developed countries, are due to overloaded networks and hot-spots. The reduction of costs in the implementation of this type of solutions is a must that should be considered before any attempt to its implementation is decided. This is one of the reasons why Tower Mounted Amplifiers, TMA, are proliferating in nowadays wireless systems. They are used in both GSM and UMTS to provide a balanced system design, allowing mobile operators to place an equal amount of receiving and transmitting sites [1]. TMAs also enable base stations to receive mobile signals more clearly in a wider coverage area than they could otherwise achieve [2]. This will allow mobile network operators to achieve the greatest possible coverage allowing less base stations and therefore limiting costs. Although TMAs emerge as an all benefit solution we must take into account the presence of interferences. We will focus specifically on the impact of nonlinear distortion in TMAs in the overall performance of GSM systems. The goal of any radio engineer when deciding to use a TMA is due to its improvement in the overall sensitivity of the system. Sensitivity gives us an indication on the robustness of a receiver in capturing a weak signal affecting directly the range of the system, and also on how immune to noise it will be. In fact if we look at expression (1), the sensitivity can be understandable as the minimum input power needed in order to get a suitable Signal to Noise Ratio (SNR) at the output of our receiver. That is the reason why sensitivity is based on the receiver noise figure, the minimum required signal-to-noise ratio for detection, and the thermal noise of the system [3], expression (1). S i, min = NF + n0 + SNR (dbm) (1) S i,min sensitivity NF noise figure of the receiver

2 SNR required output signal to noise ratio (usually related to the acceptable bit error rate) n 0 thermal noise power of the receiver, n 0 =KTB, where K Boltzman constant, T temperature and B Bandwidth of the system. Since, in expression (1) the temperature is imposed by the habitat where the TMA will be installed, B is the GSM bandwidth, 200 KHz, K is a constant, and SNR is imposed by the modulation technique, the only free parameter is NF. NF is called Noise Figure, and in a certain sense gives an idea of the SNR degradation when the signal traverses the receiver. One possible mathematical definition is expressed in equation (2). SNR i NF = (2) SNR o In a cascade of noisy blocks the overall equivalent NF is given by expression (3). NF 1 NF 1 NF i NF T = NF (3) G1 G1G2 G1G2... Gi 1 A closer look at expression (3) allows us to gather some important information. For instance the noise figure of the first block will impose the minimum noise figure of the system 1, another important conclusion is on the gain of the first components, the higher the gain the higher the desensitization of the next blocks. TMA Sensitivity Figure 1, presents a basic receiving implementation that will be used to see those different impacts of each sub-system. 1 Remember that NF is always positive and higher, or equal, to 1, by definition.

3 Figure 1 Basic Receiver NF T L 1 NF 1 inband LNA mixer = Lcable (4) Lcable Lcable Linband Lcable Linband GLNA NF 1 As can be seen in (4), cable losses are the dominant factor in the system s noise figure, thus it is the major limiting factor to receiver s sensitivity. Therefore, the idea is to desensitize the cable noise figure, and in order to achieve that manufacturers provided TMAs. A schematic TMA system is presented on figure 2. The TMA sub-system is placed near the antenna and in a certain way is similar to the well known LNB that was used for several years in the television satellite receivers; the main difference is that now no translation of frequency is needed.

4 Figure 2 TMA Schematics To evaluate the impact of introducing a TMA let us make some simple calculations. For GSM case we typically have: n 0 = -121 dbm S i,min = -105 dbm SNR = 9 db (typical value) 7 Using equation (1), NF is calculated to be 7 db ( F = = 5 ). If the system was designed in order to have a 7dB of Noise Figure, then a cable will degrade severely the overall system, and considering a cable of 3dB losses, we will have: 5 1 FT = 2 + = 10 NFT = 10dB (4) 0.5 And so the sensitivity will be: S=-102dBm. Consider now that we add a typical TMA with NF = 1.7 db and G = 12 db, applying equation (3) we will have: F = = 2 NFT = db (5) T 3 From equation (1) we can calculate our new sensitivity: S i = = 109dBm

5 This new sensitivity value allows a better reception of the signal, which can be translated into the maximum coverage area. Using the Friis equation and the Hata-Okumura propagation model [4] we reach the following maximum distances for an urban environment, considering an emitter with 30dBm of transmitted power. BTS 5.6 km MS TMA BTS 8.9 km MS Figure 3 TMA benefits The main conclusion here is that just by adding a TMA we improved our coverage by 63 %. Despite TMA benefits are important, in terms of sensitivity improvement, it can change to a very bad design decision if the presence of interferences [3] is not accounted for. In our case we will look to the specific case of the nonlinear distortion generated at the TMA itself. The Nonlinear distortion problem in TMA based sub-systems

6 Since a TMA is an active device, it will generate some form of distortion [6] that is mainly due to the finite amount of energy that can be used from the power supply. That is the reason why in some way any amplifier will always saturate for a certain amount of input power. If we make the same approach as used before when the sensitivity of a TMA was studied, then no problem will appear, since we are only dealing with small signal excursion input signals. Nevertheless if we refer now to a high power interferer, then the scenario changes, since we do not know the power of that interferer. In order to better understand the nonlinear mechanism lets approximate our amplifier with a low degree polynomial such as [5]: 2 [ x() t ] = a x() t + a x() t a x() 3 y t When, for instance, we introduce at the input of this device a two-tone signal x t) = A cos( ω t + φ ) + A cos( ω t + ) (6) ( φ2 we will have at the output various different spectral components. Despite the nonlinearity generates spectral components all over the band, in our study the most important ones are those who fall inside our bandwidth, so only two types of output spectral components will be studied, the third order IMD, and the co-channel distortion, or usually called desensitization. The IMD distortion is the one responsible for the well known spectral regrowth effect, and in a two tone excitation will appear at 2ω 1 -ω 2 and 2ω 2 -ω 1. (5)

7 Figure 4 Intermodulation Distortion and Spectral Regrowth Since in GSM there are several operators, and for each operator, due to capacity problems, there can be several emitting and receiving channels, it is quite obvious that this kind of distortion could have some impact in the performance of our system, since two different carriers will generate other two that can fall exactly over our signal. In the second case, the result is even more disastrous, since an interferer can be so strong that our signal will be destroyed [6].

8 Figure 5 - Desensitization In both cases, if the interferer signal is strong enough it will degrade our signal so drastically that it can be blocked, in fact in the World War II, this was one of the electronic war technologies, called jamming. TMA performance degradation Recalling again figure 2, where the TMA internal configuration was presented, and considering that the isolation between the Tx and Rx is high in the duplexer to prevent that the Tx signal passes throw the Rx filter and cause any nonlinear distortion, we will study the impact in our system when it receives two different Rx signals, a desired signal and an interference signal. This interference can be from the same, or another, operator, but mainly we will consider it from a different operator, because otherwise using some form of power control we could minimize the interference.

9 Mobile A GSM Desired Signal Mobile B GSM Interference Signal TMA Figure 6 TMA Receiving Two GSM Signals The first study that will be made is to calculate the values of out-of-band power needed to degrade the useful signal, considering that a 9dB SNR as to be achieved. We will use the following typical TMA values: n 0 = -121 dbm SNR = 9 db IP 3 = 25 dbm (typical value of a TMA amplifier) The system is only useful when the nonlinear distortion generated by the interference is 9dB below the sensitivity or higher, therefore, for the worst case, we have at the amplifier s output a minimum interferer signal power of: P INT = G + S SNR (7) i where P INT is the interferer power. Considering the case a of two channel interference, of equal amplitude, the third order IMD value will be: A NL = a3 A1 A2 = a3 A2 (figure 7) at 2ω 1 -ω 2 and 2ω 2 -ω 1 (co-channel interference) [6]. 8 8 Solving this two equations we reach a value of P INT =P 2 =-29 dbm. This means that if a signal at the input of the TMA reaches this value, one should expect an intermodulation power that degrades our system in a neighbour channel.

10 Figure 7 Intermodulation Power If the desensitization is calculated then it causes and amplitude interference of 3 2 A NL = a3 A1 A2 at ω 1 [6]. 4 Considering the minimum power that can be allowed by the distortion nonlinearity at ω 1 (P 1 =-105-9=-114 dbm) the interference power needed in order to generate this distortion is P INT =P 2 =6 dbm. Figure 8 Desensitization Power

11 The case before considered that the TMA had a full uplink bandwidth, that means it receives and amplifies all the GSM operators. If we now have access to a subbanded TMA, for only one operator 2, with a typical out-of-band attenuation of 80 db, this interferer powers increases to the value of 51 dbm (126 W) for intermodulation and 86 dbm (398 KW) for desensitization. And of course in this case it is unreal to consider any interference for these sources. Nevertheless and although these values seem quite high we must not forget that we might have Bit-Error-Rate (BER) degradation for lower interferer powers than sensitivity. Also we must not forget that these calculations were made assuming a two tone input. A real signal would be better modulated by a multi-tone signal or real signal [6]. Therefore, to evaluate the real impact of TMA nonlinear distortion a computer simulation was performed using a system simulator [7]. Let us consider two different operators at frequencies MHz (operator 1) and 900 MHz (operator 2). The output power of the desired signal (operator 1) is fixed at 105 dbm while the Output power of the interference signal (operator2) will be raised from -105 dbm to 30 dbm, the maximum output power of a Mobile Station. Figure 9 Simulation Schematics 2 In this case a filter should be provided, in order to attenuate the near operators.

12 BER Interference Signal Power (dbm) Figure 10 BER Curve for Desensitization We can see that the curve will quickly tend to 100 % BER. Since in the GSM case we typically allow a maximum BER of 0.2 % to guarantee receiver quality [8], we find that the maximum interferer power allowed would be approximately -14 dbm at the input of our TMA This kind of powers can easily be found in urban environment were we have a high density of BTS and so of TMAs is available, mainly in hot-spot situations as commercial malls or garages. Concluding, though TMA boosts our system performance, in the presence of strong interferer signals it can be easily blocked. In these situations we should opt for subbanded TMAs which allow us to attenuate the interferer power and so the impact of nonlinear distortions. References [1] Ira Wiesenfeld, Testing tower top amplifiers, Mobile Radio Technology, May 1, [2] LGP Telecom, Tower Mounted Amplifier Systems, Application Note [3] MAXIM, Improving Receiver Sensitivity with External LNA, APP 1836, December 27, 2002.

13 [4] Theodore S. Rappaport, Wireless Communications: Principles and Practice, Prentice Hall, New Jersey, 1996 [5] N. B. Carvalho and R. C. Madureira, Intermodulation Interference in the GSM/UMTS Bands, III Conferência de Telecomunicações, Figueira da Foz, pp , April [6] José Carlos Pedro and Nuno Borges de Carvalho, Intermodulation Distortion in Microwave and Wireless Circuits, Artech House Publishers: Norwood, August [7] Advanced Design System, 2002, Agilent Technologies [8] ETSI TS v8.9.0 ( ), Digital Cellular Telecommunications System (Phase 2+); Radio Transmission and Reception, ETSI