GB2441374A

GB2441374A - Method for optimized capacity estimation of a cellular communications network

Info

Publication number: GB2441374A
Application number: GB0617245A
Authority: GB
Inventors: Nico Lugil; Wouter De Win
Original assignee: Agilent Technologies Inc
Current assignee: Agilent Technologies Inc
Priority date: 2006-09-01
Filing date: 2006-09-01
Publication date: 2008-03-05
Also published as: GB0617245D0

Abstract

A method for estimating the capacity of a cellular communications network. The methods comprising modeling an inner cluster of cells by applying random statistics, modeling an outer cluster of cells by applying a relative interference factor, calculating a total interference towards a center cell of the inner cluster, and using the calculated total interference to derive a number of mobile terminal users in the center cell.

Description

METHOD FOR OPTIMIZED CAPACITY ESTIMATION OF A CELLULAR

COMMUNICATIONS NETWORK

Field of the invention

The present invention relates to a method for optimized capacity estimation. More particularly, the present invention discloses a method for capacity estimation of the reverse link in a cellular system in which the user terminals make use of transmit power control.

State of the art By its nature, the reverse link capacity of CDMA communication systems is interference limited. Although the interference signal power is attenuated to a negligible value beyond the radio horizon, the impact of interference still propagates indirectly over the whole network because of the power control mechanism. As a result, every user has a direct or indirect impact on the amount of interference seen by every other user in the network. Hence, to estimate the capacity, one should calculate the transmit power of every user for the network configuration of interest.

Although this is mathematically possible, the practical implementation becomes unmanageably complex for a realistic number of cells.

To limit the complexity, currently applied methods for capacity estimations are based on a simplified representation of the network and hence lead to rather rough estimates. Figure 1 is an example of a classical approach for capacity estimations of CDMA communication systems and is based on a virtual network 10 that is limited tc one cell 1, where interference from the surrounding cells 2, each having a plurality active users U, are represented by an interference factor relative to the intracell interference power. Although this approach includes indirectly an infinite number of cells, it does not obey the statistical randomness found in a realistic network.

Another known method based on the Erlang capacity does involve statistic elements. However this method still assumes that the interfering cells are identically populated with mobile stations, more specific with respect to the number of mobile stations, their locations and their signal-to-noise requirements. Hence, the behavior of the interfering cells is correlated with themselves and the cell of interest.

Aims of the invention The present invention aims to provide a method for accurate estimation of the capacity of the reverse link in a cellular system exploiting transmit power control for its terminals, such as UMTS/FDD. The proposed method overcomes the drawbacks of the above mentioned prior art solutions.

According to the present invention, that aim is achieved by means of a method having the features set forth in the claims that follow, such claims being an integral part of

the present disclosure.

The invention has possible applications in the following

fields (non-limitative list)

* Capacity estimation related to cellular network planning; * Cellular network monitoring; * Load testing for cellular base stations; * Realistic channel emulation including other users interference generation.

* Evaluation of air-interface design with respect to impact on interference and capacity.

Short description of the drawings

Fig. 1 represents a drawing of the prior art capacity estimation: other-cell interference is represented by a relative interference factor based on virtual copies of the centre cell.

Fig. 2 represents the method of the present application, by involving random variables in a number of surrounding cells and by applying multiple independent interference factors to model the outer cells.

Acronyms BS Base Station MS Mobile Station SNR Signal to Noise Ratio Q0S Quality of Service

Detailed description of the invention

The proposed method in this text gives a solution that combines a manageable complexity with a statistical correct representation of the communication network. Supplementary, the method is easily scalable so that improved accuracy can be obtained when more processing power and/or tune is available.

The present invention, as shown in figure 2, is based on a split of the network 20 into two groups of cells. The first group consists of a "centre cell" 21 and a limited amount of surrounding rings 22, 23... together called the "inner cluster" 24. The second group, called the "outer cluster" 25, consists of all the cells surrounding the inner cluster and is in essential unlimited in size. In addition, each cell has an antenna A and a plurality of active mobile terminal users U. Within the inner cluster, the network configuration is obeying random statistics. Random variables are, for instance, the number of users per cell, their position U and their QoS requirements. Based on the power control mechanism, transmit powers and their corresponding interference effects on each other are calculated for users within the inner cluster. To include the effect of interference from the users that reside in the outer cluster, a specific relative interference factor for every cell in the inner cluster is calculated. This set of relative interference factors is taken into account when performing the power control for the users within the inner cluster.

Although the relative interference factors inherently introduce correlation between interference from the outer cluster and intra-cell interference for the targeted cell within the inner cluster, the effect on the overall capacity calculation is randomized over all of the cells within the inner cluster. The degree of randomization increases with increasing number of cells within the inner cluster resulting in increased accuracy. Because the solution converges rather quickly towards the theoretical solution, a limited amount of cells within the inner cluster will already provide an acceptable accuracy.

The power control calculations are as follows.

Because the new capacity estimation method simulates a substantial part of the whole network rather than just a single cell, an accurate power control mechanism is required in order to calculate the individual transmit powers for every mobile station residing in the "subnetwork" of the inner cluster. The knowledge of the individual transmit powers is an important element of the new method because it allows a realistic estimation of the corresponding interference powers. Here we illustrate how these power control calculations are performed.

We define the tolerated interference over signal power matrix R0j as SJNR' 0 0 R = 0 SJNR2 0 with SINR the required signal-to-noise ratio of the nth MS N the total number of MSs in all cells 0 0 0 SINRN' and the channel power weighting matrix W as 0 0 wth = + traffic channel 2 power'1 keep -alive channel power -0 w;1 " traffic channel I power) traffic channel I power 0 fa = I for MSs with an active traffic channel where 1; = 0 for MS's with an inactive traffic channel and the received reverse link power vector Prx as p1.1 p2.' PrX = with the received reverse link power of MS u located in cell c Pj,c P2c PUt:. C The tolerated interference power vector toi that a ES may receive for every MS can now be expressed as = R,1Wp We also define the generated interference over signal power matrix Rgen as 12 V V (j2 (,c \P (IC V (1C \ 1,2 I 2.2 jJ l. C I i 2.C I J ir'I I jI TJ T' Ijl L2.C) ,, 2) 2.2) tJ,2 l,C (,2 V (, "P (12 V (1c (,c V (1C \P 2J 2.2 I I &..2 I I l.C I 2,C (ICC li " Ill \ 2.2) U2) IC) L 2.C1 U.C) * , *,* V "P " (2 y I1c V (,c V (,c U 2,2 I I U 2 I. CJ I 2C I t+.c I i j +1; i+j' j, -j--j *.

. l.2, 2.2, t,, U1.2) l.C, 2.Cj U.C) (c \P (/1 \P (i V -(1c V (1c V (,c V I.c I I 2,C U.C I J 12 1+12 ** f2 jr'j rj II) 2.1) UI t,, IC) 2,C) , u.c) (i \" (i " hi Y' (1c \P (jc y (,c \P I i.c I I 2,C Uc,C I 1 1,1 2,1 I /2) l+f2 12 J2.*. I " R, II 2,1) U.l: . i. 2.C) uC f,i V (i V (i V (,c V (,c V (1C V I l.C I I 2,C I I III I 211 * I....L...J If2 1 f2 f2 12 zI (,21 Ij2 2J (2,2 1,11 2,1) I,,, U,lj _______________________________ - ic) 2,C,, \, uc) (i \P (p \P (1 (11 V (12 y (p V J-J. J.. f i+f 1,1) ,, 2,1 I u,i,s i,,i t, u) V V V 112 (j2 V 1,1 I 2.1 I. 1,2 I 2,2 I U2 I7H Ii IIc I 7'*) 171 I " I+f( f *** I+f.

1.1) 2.1, (J.1) 1.2,, I 2,21 ,, t/..21 * : , V V (i \P V (a \" (12 V 1.1 2,1 I U,,I l j 1.2 I 2,2 I U,,2 Ij IjEH l+f l f1 1+f Ic , 1.1 / I.,, 2,1) (II 1.2) ,, 2.2) 1.12,2) I J the other cell to own cell interference factor for cell c with 1,, the line-of-sight path length between MS u in cell c and the BS of cell c p the propagation path loss power coefficient The generated interference + noise power vector gen that a S BS may receive for every MS can now be expressed as 1gen = RgenPrx + with n the thermal noise power. The power control mechanism will control MS's transmit power so that 1gen 1tol and we can write Rgen Pr + = R10, WPrx That can be solved for Px-x = (R,01w -Ren)' tth The total received wideband power for the first BS can be calculated as = p11 + SINRw1' p11 and the noise rise for the same BS as rxiot noise rise = th It is not intended that the present invention be limited to the above embodiments and other modifications and variations to the method are envisaged within the scope of the claims.

Claims

Claims 1. A method for estimating the capacity of a celluar

communications network, the methods comprising the steps of: modelling an inner cluster of cells by applying randon statistics, modelling an outer cluster of cells by applying a relative interference factor, calculating a total interference towards a center cell of the inner cluster, and using the calculated total interference to derive a number of mobile terminal users in the center cell.

2. A method as claimed in claim 1, wherein the randon statistics comprises a number of users per cell and/or user position and/or user quality of service.