# Impact of Path Loss and Delay Spread on Base Station Cooperation

←

→

**Page content transcription**

If your browser does not render page correctly, please read the page content below

Impact of Path Loss and Delay Spread on Base Station Cooperation Konstantinos Manolakis, Stephan Jaeckel, Eva Salvador Marquez, Volker Jungnickel Fraunhofer Heinrich Hertz Institute Einsteinufer 37, D-10587, Berlin, Germany Konstantinos.Manolakis@hhi.fraunhofer.de Abstract—In this paper, we investigate the maximum inter-site symbol interference (ISI) between successive OFDM symbols. distance (ISD) for performing joint signal processing between There is previous work about the dependence of channel cooperative base stations. As a metric, we use the maximum delay spread from the path gain and thus the propagation excess delay measured at 95% point of the cumative power delay profile from all base stations. For the distance-dependent channel distance. In [3] it was found that the delay spread as well as parameters, we consider Greenstein’s statistical propagation it’s standard deviation increase with distance and deep shadow model, which we extended for broadband transmissions. We ex- fading in urban scenarios. An overview of the state of the art tract all model parameters from 2.6 GHz multi-cell measurements in cooperative MIMO channel modeling can be found in [4], in our field trial and parametrize the model at a fixed ISD. We where new challenges and open issues are addressed. also investigate the impact of antenna downtilt and find that when a larger downtilt is used, the rms delay spread and 95% In this paper, we consider the impact of base station (BS) excess delay are smaller. However, there are critical 3D effects cooperation on the overall delay statistics in a realistic deploy- close to the sites not included in the model. Then we consider ment scenario. We seek for an answer to the question what larger ISDs and indicate how the delay parameters grow. Based distance is allowed between cooperative base stations without on Greensteins model, the short cyclic prefix in LTE is hardly violating a given CP. As a starting point, we use a distance- violated for realistic ISD at 2.6 GHz. dependent statistical propagation model originally introduced by Greenstein et al. [5], which describes also the correlation I. I NTRODUCTION between path gain and delay spread. We parametrize this Base station cooperation is envisioned as a promising tech- model based on coherent multi-cell channel measurements nique for future mobile networks where the carrier frequency from our LTE-Advanced field trial in Berlin [6–8]. Therefore, shall be fully reused. It reduces the mutual interference we extract the distance dependencies of path loss, shadow between adjacent radio cells and increases significantly the fading and delay spread as well as their correlation at a carrier spectral efficiency [1]. In the downlink, multiple antennas at frequency of 2.6 GHz in an urban macro-cell scenario. multiple base stations are considered as inputs and multiple We illustrate that Greenstein’s model predicts the delay terminals served in the corresponding cells are considered statistics properly at least at larger distances from the BS. as outputs of a distributed multiple-input multiple-output Close to the BS, however, we observe significant deviations (MIMO) system. Recent research has revealed that spatial from the predicted statistics if the downtilt is set so that multiplexing of multiple data streams is possible provided the main beam touches the ground at 0.9 times the inter that the channel information between all transmit and receive site distance (ISD). With a downtilt of 0.33 times the ISD, antennas is provided and joint signal processing is performed as recommended by 3GPP for LTE-Advanced (see [9]), the at least at one side of the MIMO link. overall delay statistics is mostly due to the overlap region Multi-carrier transmission using orthogonal frequency divi- covered jointly between adjacent sites, where Greenstein’s sion multiplexing (OFDM) is a favorable scheme to manage model is more appropriate. Thus we can increase the ISD the complexity of the required signal processing for MIMO and predict the delay statistics by using our parametrized and is standardized in the 3rd generation partnership project propagation model. Our results indicate that BS cooperation in (3GPP) long term evolution (LTE) [2]. OFDM is well suited LTE-Advanced is feasible for distances up to 1.7 km between for transmission over frequency-selective channels, as it di- the sites at 2.6 GHz without violating the short CP. vides the available bandwidth in a number of orthogonal sub- The paper is organized as follows. In Section II, Green- channels, where each of them observes frequency-flat channel stein’s model is reviewed and extended for broadband multi- fading and can be processed separately. However, the cyclic cell transmissions. In Section III, our measurement setup and prefix (CP) needed for sub-carrier based equalization reduces the parameter extraction routines are described. Section IV the spectrum efficiency. It is chosen little longer than the presents the extracted parameters and discusses the scope of largest multipath delay in the targeted propagation environ- the model by comparing prediction and measurement results. ment. If multiple distributed base stations serve a terminal Furthermore, the parametrized model is used for predicting jointly, there are additional contributions to the overall delay the delay statistics in larger cells. Finally, conclusions are statistics which might violate the CP length and create inter- discussed in Section V.

II. T HE R ADIO C HANNEL M ODEL accepted for cellular radio channels [10]. Consequently, the A. Single-Link Narrow-Band Channel Model received power p at distance d can be expressed by multiplying the path gain g (including shadow fading variations) with a Greenstein [5] states three fundamental conjectures about unitary PDP pu according to the channel delay spread: g −τ t • the rms delay spread is lognormal at any distance p(d, t) = g pu (τrms , t) = e rms , (4) τrms • it’s median value increases with distance • the delay spread tends to grow in deep shadow fading where g and τrms are random variates that depend on the The last conjecture is very important as it suggests that distance d and are given by (1) to (3) and t denotes the dispersion is larger in case of deep shadow fading, when the time-delay domain. The PDP of the received signals that are signal to noise ratio (SNR) is already weak. In [5] physical radiated from Nt distributed transmit antennas is expressed by arguments that support these conjectures are explained and the sum of the PDP of the single links pi : confirmed by the existing literature. Finally, a mathematical Nt X Nt X gi t−ti −τ model for the joint statistics of average path gain and rms P (d, t) = pi (di , t − ti ) = e rms,i . (5) τrms,i delay spread is proposed. Parameters are specified for various i=1 i=1 environments in [5]. The time-shift ti = di /c to the ith transmit antenna is The path gain is given by: calculated with respect to the closest base station, which means g = G1 d−γ x , (1) that the first path that is received by the mobile user at the time t = 0 arrives from the closest base station; c is the speed where d is distance in kilometers, G1 is the median value of of light. path gain g at a reference distance of d = 1 km, γ is the path loss exponent (usually between 2.5 and 4) and x is a III. M ULTI -C ELL M EASUREMENTS lognormal variate at distance d. Specifically, X = 10log(x) We have set up a multi-cell testbed with 3 multi-sector sites is a Gaussian random variable with zero mean and a standard in the city center of Berlin, Germany. These BSs are placed on deviation σx . top of the Heinrich Hertz Institute (HHI), the main building of The median value of the rms delay spread τrms increases the Technical University of Berlin (TUB) and the Telefunken according to distance with an exponential factor ǫ. The rms Tower Deutsche Telekom Laboratories (TLabs) as seen in delay spread at distance d is given by Fig. 1. Detailed information on this system are available in our publications [7, 11, 12]. A precise time reference is available τrms = T1 dǫ y , (2) using an advanced global positioning system (GPS) receiver where T1 is the median value of τrms at d = 1 km, the at each site and the whole radio network is synchronized to exponent ǫ lies between 0.5 and 1.0, and y is a lognormal the one pulse per second (1PPS) signal. In conjunction with variable. Specifically, Y = 10log(y) is a Gaussian random our coordinated multi-point (CoMP) experiments, so called variable at distance d, which has zero mean and a standard channel state information (CSI) reference signals were defined deviation σy . to identify the channels to each of the BSs. Each BS is The delay spread is correlated by a negative correlation identified by a pilot comb in frequency domain and we apply a coefficient with the shadow fading gain. If strong channel taps, cyclic shift of the comb along the frequency axis by an integer which usually arrive at early time instants, are blocked by large number of sub-carriers where the shift identifies the cell. objects (negative shadowing gain), τrms is expected to grow. For identifying multiple antennas in each cell, we have used Hence, the correlation between g and τrms is described by four consecutive OFDM symbols where another orthogonal sequence is transmitted for each antenna [13]. In this way, we E(XY ) = ρ σx σy . (3) have realized a distributed, but fully phase-coherent channel Here, ρ is the correlation coefficient between the two Gaussian sounder for up to 6 cells. variables X and Y and E(·) denotes the expectation operator. We use +18 dBi BS antennas from Kathrein with two cross- According to the model, large τrms are associated with smaller polarized ports ±45◦ at the transmitter. The downtilt for all g values, ρ should thus be negative; a value of ρ = −0.75 was BSs was first set such that the beam hits the ground at a proposed in [5] for all types of environments. distance of 450 m, which corresponds to 0.9 times the ISD. In a second measurement, the downtilt was set so that the B. Multi-Link Broadband Channel Model beam hit the ground at 0.33 times the ISD, as explained in The channel model of Section II-A was initially introduced Section I. Measured values for the 95% excess delay are shown for narrow-band radio channels, where the sampling time is on the maps in Fig. 1 and 2 for both downtilts. Additional in the same order of magnitude or larger than τrms . This measurement parameters are available in Table I. assumption is not valid for broadband systems like LTE, where At the receiver side, we use our test terminal which is the sampling time is significantly shorter. Therefore, we have synchronized over the air. The mobile terminal (MT) detects to introduce a power delay profile (PDP). Our assumption the CSI reference signals (CRS) and converts them into an is that the PDP follows an exponential decay, which is well Ethernet packet stream as described in [12]. This packet data

N 0.95 Excess Delay 3 The import filter extracts a 2 × 12 × 144 channel tensor for HHI Multicell @ 0.9 ISD 2.7 each 10 ms of the recorded packet stream. The dimensions height: 64 m tilt: 8° 2.4 correspond to the number of receive antennas, the number of Gu 1 2.1 transmit antennas and the number of pilots, respectively. The eric kes 3 tr . 1.8 144 channel coefficients of each Tx-Rx link can be seen as 2 0.95 Excess Delay [us] . str yn = hn + vn , n ∈ {1, 2, ...144} rch 1.5 (6) Ma Fr a unh o Sa lz 1.2 fer uf str . er where y is the observation (or measurement) of the broadband Ei ns te 0.9 in u radio channel h in frequency domain at N fixed pilot positions fe r TLabs height: 84 m 0.6 tilt: 10° νn . These pilot positions are not equidistant in our system. 0.3 1 Er n st-R Pla eute v is the estimation error, which is assumed to be additive 3 tz r Juni 0 2 Str. des 17. 1 TUB white Gaussian noise (AWGN). We now use the preprocessing 3 height: 49 m tilt: 6° technique described in [8, 14] to extract the contributing mul- 2 tipath components. Essentially, the preprocessing calculates all 500m parameters α and φ and τ which sum up to L Fig. 1. Multi-cell 95% excess delay measurements. The main beam of X antennas hits the ground at 0.9 times the inter-site-distance. h̃n = αl · ejφl · e−2πj·τl ·B·νn (7) l=1 N 0.95 Excess Delay 3 where αl is the amplitude, φl is the phase and τl is the delay HHI Multicell @ 0.33 ISD 2.7 of each multipath component. These parameters are the input height: 64 m tilt: 21° 2.4 for all following evaluation steps. Due to the processing, h̃ has 2.1 a SNR which is approximately 6 dB better than in y since a Gu eric 3 1 kes tr . 1.8 significant part of the noise can be removed. 2 We calculate the shadow fading according to T.S. Rappaport 0.95 Excess Delay [us] . str rch 1.5 Ma Fr a [10]. The instantaneous path gain is first estimated from the unh o Sa lz 1.2 fer str . uf er preprocessed data by averaging all signal components over all Ei ns te 0.9 in transmit and receive antennas of one sector u fe r TLabs height: 84 m 0.6 tilt: 27° nt Xnr L 1 X X 1 Er n st-R 0.3 g= · Pr,t Pr,t = α2l,r,t (8) 3 Pla eute tz r Juni 0 nt nr t=1 r=1 2 Str. des 17. l=1 3 1 TUB height: 49 m tilt: 17° where g is the total power of a PDP. Power values for the 2 500m two sectors of each site are averaged. The resulting values are then additionally averaged over segments of 5 m length to remove small scale fading effects which results in a total Fig. 2. Multi-cell 95% excess delay measurements. The main beam of antennas hits the ground at 0.33 times the inter-site-distance. of 692 measurement values per BS. The shadow fading (1) in logarithmic notation notes g dB (d) = GdB 1 + γ · 10 log10 (d) + X . (9) stream is tapped at the terminal and recorded using a notebook computer. The terminal performs coarse timing adjustment We calculate GdB 1 and γ by linearly fitting the measured data with respect to the OFDM CP. As a result, the mean multi- to the linear slope and then calculating the remaining variance. cell delay is measured and the impulse response is placed The rms delay spreads are calculated by according to the CP. v u L uX α2 L l X α2l τrms = t · τl2 − τm 2 τm = · τl (10) Pt Pt TABLE I l=1 l=1 M EASUREMENT PARAMETERS with τl as the delay- and α2l the power of the lth tap. τm is the Parameter Value mean delay and Pt is the total power. Again, we calculated Center frequency 2.68 GHz the parameters T1 and ǫ by transforming (2) into logarithmic Bandwidth (B) 18.36 MHz No. of pilots (N ) 144 domain and linear fitting. MIMO capabilities 2x2 per BS At that point it has to be mentioned, that measurements from No. of BSs 6 (up to 12x2 links) positions closer than 210 m to the BS were excluded from the CSI update interval 10 ms Maximal speed 2.8 m/s ≈ 10 km/h evaluation. At these positions large propagation delays due to Inter site distance ≈ 500 m multiple reflexions on surrounding buildings were observed, BS Tx power 36.5 dBm while the direct signal was attenuated by the Tx antenna MT noise floor -95 dBm pattern. Such critical effects could only be captured by a 3D

TABLE II PARAMETERS E STIMATED FROM M EASUREMENT parameters listed in Table II. Evaluation of τrms is performed over 5000 statistically independent realizations. The τrms is Parameter Estimated Value (HHI/TUB/TLabs) calculated according to (10). The model prediction leads to Average path gain (G1 ) 107.4 / 126.9 / 106.6 [dB] the solid red line in Fig. 3, which is actually not far from the Path loss exponent (γ) 2.0 / 6.3 / 3.0 Median of τrms (T1 ) 0.4 / 0.3 / 0.3 [µs] measured (dashed) line. Exponent ǫ 0.5 / 0.2 / 0.2 Fig. 4 illustrates the CDF of the τ95 , i.e. the 95-percentile of Variance σx 5.5 / 7.8 / 6.9 [dB] the cumulative PDP, that limits the signal to interference and Variance σy 1.7 / 1.8 / 1.5 [dB] Correlation coefficient (ρ) -0.7 / -0.3 / -0.6 noise ratio (SINR) contribution of the ISI to 13 dB. For the same measurement and simulation scenarios like τrms , results run into similar observations. The model predicts the expected delay parameters quite precisely, only slightly more optimistic. channel model and were thus excluded here. All parameters estimated from measurements are listed in Table II. IV. S TATISTICAL E VALUATION AND R ESULTS 1 First we validate our propagation model by parameterizing 0.9 it and comparing predictions with measurements. We study 0.8 the impact of antenna downtilt and look into the cases where 0.7 the main beams hit the ground at 0.9 and 0.33 times the ISD. 0.6 Finally, we increase the distance between the base stations for P[τ ≤ x] 0.5 95 predicting the delay statistics according to our model. Fig. 3 illustrates the cumulative distribution function (CDF) 0.4 of the τrms for the multi-cell channel. The black curve shows 0.3 the measured values for 0.9 times ISD antenna downtilt, while 0.2 measured 0.90 ISD measured 0.33 ISD the blue curve for 0.33 times ISD. These values were taken measured 0.90 ISD, d ≥ 210m 0.1 from the route shown in Fig. 1. It is observed that by deploying modeled 0.90 ISD, d ≥ 210m 0 a larger downtilt, significantly smaller rms delay values can be 0 0.5 1 1.5 2 2.5 3 τ95 [µs] realized. The illuminated region in the cell is then more close to the BS and scattered signals have less delay. The statistics shown by the dashed red line are mostly due to the overlap Fig. 4. Multi-cell 95% excess delay τ95 . The main beam of antennas hits the ground at 0.9 or 0.33 times the inter-site-distance. Model predicts well region covered jointly between adjacent sites. for positions at least 210 m away from the base stations. 1 Finally, we use the model for predicting the delay statistics 0.9 for larger ISD (dISD ). Channel parameters from Table II and uniform user allocation is assumed within the region defined 0.8 by the three BS. Fig. 5 and Fig. 6 illustrate the τrms and τ95 0.7 respectively for dISD = 500, 1000 and 1732 m (maximum 0.6 distance according to LTE [2]). It can be observed, that even P[τrms≤ x] 0.5 in for the maximum ISD, τ95 exceeds the short CP length of 4.7 µs in LTE in less than 5% of the cases. 0.4 0.3 V. C ONCLUSIONS AND D ISCUSSION 0.2 measured 0.90 ISD We performed multi-cell broadband channel measurements measured 0.33 ISD 0.1 measured 0.90 ISD, d ≥ 210m in our testbed and studied the impact of antenna downtilting modeled 0.90 ISD, d ≥ 210m on the channel delay statistics. It was found that when using a 0 0 0.2 0.4 0.6 0.8 1 larger downtilt, significantly smaller channel delay spreads can τ [µs] rms be realized. We extended Greenstein’s statistical propagation model for covering broadband transmission from distributed Fig. 3. Multi-cell rms delay spread τrms . The main beam of antennas hits the ground at 0.9 or 0.33 times the inter-site-distance. Model predicts well base stations. Based on measured data from our field trial for positions at least 210 m away from the base stations. we parametrized the model and compared measurements with simulations. The model was verified in terms of rms delay In the next step, a multi-cell simulation environment with spread and 95% excess delay. Despite of using a 2D model an ISD dISD =500 m is used an MT are placed on the in a 3D setup, it predicts the delay parameters in the multi- same positions as in the measurement, as shown in Fig. 1. cell case quite precisely at least at larger distances from the Independent channels are generated for the HHI, TUB and base stations. Close to the base stations, however, we observe TLabs links by using the channel model as in (1) to (3) and deviations from the predicted statistics if the downtilt is set

1 ACKNOWLEDGMENT 0.9 The authors would like to thank the German Ministry d =500m ISD of Education and Research (BMBF) for funding this work 0.8 dISD=1000m dISD=1732m through the national research project EASY-C (Enablers of 0.7 Ambient Services and Systems Part C – Wide Area Coverage). 0.6 P[τrms≤ x] R EFERENCES 0.5 [1] M. Karakayali, G. Foschini, and R. Valenzuela, “Network 0.4 coordination for spectrally efficient communications in cellular 0.3 systems,” IEEE Wireless Commun. Mag., vol. 13, pp. 56–61, 2006. 0.2 [2] 3GPP LTE, E-UTRA and E-UTRAN – Overall description 0.1 (Release 10), Std., June 2010. [3] E. Sousa, V. Jovanovic, and C. Daigneault, “Delay spread 0 measurements for the digital cellular channel in toronto,” IEEE 0 0.5 1 1.5 2 2.5 τrms [µs] Trans. Veh. Technol., vol. 43, no. 4, pp. 837–847, 1994. [4] C.-X. Wang, X. Hong, X. Ge, X. Cheng, G. Zhang, and Fig. 5. Predicted multi-cell rms delay spread τrms for larger inter-site- J. Thompson, “Cooperative mimo channel models: A survey,” distances and uniform user allocation between the base stations. Communications Magazine, IEEE, vol. 48, no. 2, pp. 80 –87, 2010. [5] L. Greenstein, V. Erceg, Y. Yeh, and M. Clark, “A new path- 1 gain/delay-spread propagation model for digital cellular chan- nels,” IEEE Trans. Veh. Technol., vol. 46, no. 2, pp. 477–485, 0.9 1997. 0.8 d =500m [6] S. Jaeckel, L. Thiele, A. Brylka, L. Jiang, and V. Jungnickel, ISD 0.7 “Intercell interference measured in urban areas,” Proc. IEEE d =1000m ISD ICC ’09, 2009. d =1732m 0.6 ISD [7] V. Jungnickel, M. Schellmann, L. Thiele, T. Wirth, T. Haustein, P[τ ≤ x] 0.5 O. Koch, E. Zirwas, and E. Schulz, “Interference aware schedul- 95 ing in the multiuser MIMO-OFDM downlink,” IEEE Commun. 0.4 Mag., vol. 47, pp. 56–66, 2009. 0.3 [8] S. Jaeckel, L. Thiele, and V. Jungnickel, “Interference limited MIMO measurements,” Proc. IEEE VTC ’10 Spring, 2010. 0.2 [9] L. Thiele, T. Wirth, K. Brner, M. Olbrich, V. Jungnickel, 0.1 J. Rumold, and S. Fritze, “Modeling of 3D field patterns of downtilted antennas and their impact on cellular systems,” Proc. 0 0 1 2 3 4 5 6 WSA ’09, 2009. τ95 [µs] [10] T. Rappaport, Wireless Communications. Principles and Prac- tice, 2nd ed. Prentice Hall, 2002. Fig. 6. Predicted multi-cell 95% excess delay τ95 for larger inter-site- [11] V. Jungnickel, L. Thiele, T. Wirth, et al., “Coordinated multi- distances and uniform user allocation between the base stations. point trials in the downlink,” Proc. IEEE Globecom Workshops ’09, 2009. [12] V. Jungnickel, A. Forck, S. Jaeckel, F. Bauermeister, S. Schif- in a way that the main beam touches the ground at 0.9 fermueller, S. Schubert, S. Wahls, L. Thiele, T. Haustein, W. Kreher, J. Mueller, and H. D. G. Kadel, “Field trials using times the inter-site-distance (small downtilt). With a larger coordinated multi-point transmission in the downlink,” Proc. downtilt set to 0.33 times the ISD, the overall delay statistics IEEE PIMRC ’10 WDN-Workshop, 2010. is mostly due to the overlap region covered jointly between [13] V. Jungnickel, K. Manolakis, L. Thiele, T. Wirth, and adjacent sites, where the model is more appropriate. Finally T. Haustein, “Handover sequences for interference-aware trans- we consider larger ISDs and predict the delay statistics by mission in multicell MIMO networks,” Proc. WSA ’09, feb 2009. using our parametrized propagation model. Our results indicate [14] S. Jaeckel and V. Jungnickel, “Estimating MIMO capacities that base station cooperation in LTE-Advanced is feasible for from broadband measurements in a cellular network,” Proc. distances up to 1.7 km between the sites at 2.6 GHz without EUCAP ’10, 2010. violating the short cyclic prefix.

You can also read