EP1198964A1

EP1198964A1 - Method and system for dynamic allocation of radio channels in digital telecommunication networks

Info

Publication number: EP1198964A1
Application number: EP00954564A
Authority: EP
Inventors: Fulvio Margherita; Sergio Parolari
Original assignee: Siemens Information and Communication Networks SpA; Siemens Information and Communication Networks Inc
Current assignee: Siemens Information and Communication Networks SpA; Siemens Communications Inc
Priority date: 1999-07-30
Filing date: 2000-07-24
Publication date: 2002-04-24
Also published as: CA2391689A1; WO2001010155A1; ITMI991710A0; IT1313314B1; ITMI991710A1

Abstract

Method for the dynamic allocation of radio channels (Ci) in digital telecommunication networks with time division duplex access, whose radio signals are divided into frames having pre-determined duration and each frame is subdivided into a pre-determined number of time intervals (Ti) which are assigned priority values (Pi) based on measures of channel interference and/or quality (Ci), each communication service (Sx) employing a particular number (Rx) of said channels (Ci) at a time. This method includes at least a measurement of the signal attenuation (PLx) with which said communication service (Sx) has been requested, as well as the allocation of said number (Rx) of channels (Ci) of the communication service (Sx) in a time interval (Tx) having an increasing priority value (Pi) with the attenuation (PLx) of the relevant signal, in order that the services employing said number (Rx) of channels (Ci) are allocated in time intervals (Ti) having increasing priority values (Pi) with the attenuation of the relevant signal.

Description

METHOD AND SYSTEM FOR DYNAMIC ALLOCATION OF RADIO CHANNEL IN DIGITAL TELECOMMUNICATION NETWORKS

5 Field of the Invention

The present invention relates to a method for the dynamic allocation of radio channels in digital telecommunication networks, in particular with time division duplex access or TDD (Time Division Duplex), such as for instance mobile telecommunication networks belonging to DECT or UTMS-TDD standards. The present invention relates

10 also to a system implementing this method.

It is well known that in mobile telecommunication networks with TDD access, the transmission and reception of radio signals from and to the base stations do not occur at the same time, but are alternated in a continuous sequence of periods having predetermined duration, each of them called frame and opportunely coded and identified

15 by the system. In particular, each frame is divided into a pre-determined number of time intervals or timeslots, they too having pre-determined duration, part of which is destined to transmission and part to the reception of the signals from base station to user equipment. Each one of these timeslots can also be subdivided into a plurality of codes representing the elementary resources (channels) assigned in the 0 communication.

At each communication service between a mobile unit and a base station one or more channels of a particular time slot are generally assigned, which contains at most Nmax channels, according to the requested transmission speed. Background art

25 The selection of said timeslot is made through a dynamic channel allocation procedure based on priority values Pi calculated for each timeslot. In this procedure disclosed in the article by Y. Furuya and Y. Akaiwa under the title "Channel Segregation, A distributed Adaptive Channel Allocation Scheme for Mobile Communication Systems", Second Nordic Seminar on digital Land Mobile Radio

30 Communication, 14-16 October 1986", pp 311-315, the control processor of the base station performs at each service request, a calculation of the priority values Pi on the basis of interference and/or quality measures of the channels, so that the timeslots available for the allocation of channels result only those whose priority value is higher than a given pre-set threshold value Pt. The calculation of priority values Pi of each timeslot after k service requests is generally made through the following iterative formula: where k is the number of connection service requests from the moment of system starting, Ns,(k) is the number of successful connections and s,(k) is a logic function returning 0 or 1 on the basis of the negative or positive result of the connection, respectively.

Observing such formula, it can be noticed that at starting, with small k values, the calculation of priority values very quickly adapts to the network characteristics, but results slowed when k values increase, therefore the above mentioned method known for the allocation of channels shows a high risk for connection losses in case the network traffic distribution suddenly changes, for instance when number of connection service requests occur, concentrated in time. Summary and scope of the Invention Object of the present invention is therefore that to give a method for the dynamic allocation of channels which is free from this drawback, as well as a system implementing this method. Said object is attained with a method and a system whose main characteristics are specified in claims 1 and 12, respectively, while additional characteristics which are believed to be novel are specified in the appended claims. The method according to the present invention results much more rapid than the known methods, since a partial re-ordering of channels allocated in the different timeslots is made, that is, at each service requests, only the services employing the same number of channels of the requested service are re-ordered.

Moreover, the quality of channels allocated through the method according to the present invention is generally bettered compared to that of the channels allocated through the known methods. In fact, at equal number of allocated channels, the services with signals having high attenuation values or pathloss are allocated in timeslots having high priority values, so that the allocated channels can be shared in the different timeslots in the best way according to the quality of the relevant signals. Another advantage of the method according to the present invention is represented by the use of a new kind of formula for the calculation of priority values which, contrarily to the above mentioned formula of the known type, enables to discretionary adjust the system adaptation speed to the contingent situation of the network traffic, that is to the interference and/or quality variations of the channels.

A further advantage of the method according to the present invention is represented by the fact that said allocation and release algorithms can be structured in such a way to give preference, if necessary, to the services employing a low or high number of channels.

Brief description of figures

The present invention together with further advantages and characteristics thereof may be understood by those skilled in the art making reference to the following detailed description taken in conjunction with the accompanying drawings in which:

- figure 1 shows a partial block diagram of a system implementing the method according to the present invention;

- figure 2 shows a flow chart of an allocation algorithm of an embodiment of the method according to the present invention; and - figure 3 shows a flow chart of a release algorithm of an embodiment of the method according to the present invention.

Detailed description of a preferred embodiment of the Invention Making reference to figure 1 , it can be noticed that a system implementing the method according to the present invention includes in a known way, a plurality of base stations 1 belonging to a digital telecommunication network with time division duplex access, such as for instance a mobile telecommunication network belonging to the UTMS standard, which communicate through radio signals with a plurality of user equipment 2. One or more channels Ci of a timeslot Ti are generally assigned to each communication service Si made by base stations 1 (only 8 timeslots Ti of communications originated by the user equipment 2 are shown in the figure, for representation simplicity). Furthermore, a univocal priority value Pi is assigned to each timeslot Ti which, however, can vary in time according to the result of a known formula of the type described above or of a new formula which shall be described here after. Said priority values Pi are based on interference and/or quality measures of communication channels Ci between base stations 1 and user equipment 2.

According to a preferred embodiment of the invention the interference and/or quality measures of channels are made measuring the "path loss", that is the attenuation of the signal transmitted by the user equipment 2. According to the invention, the communications with higher path loss are allocated in timeslots with higher priority Pi, that is in channels capable of ensuring a better transmission quality. On the contrary, communications with lower "path loss" are allocated in timeslots with lower priority Pi, that is in channels capable of ensuring a lower quality.

Tentatively said situation has been represented in figure 1 where for each station 1 a first coverage area 3 and a second coverage area 4 are represented: In figure 1 it is assumed that users located in the coverage area 4 are more distant from the relevant base station 1 and therefore communications shall be reasonably characterised by a higher "path loss" (they shall therefore be assigned a timeslot having higher priority Pi), while the users in coverage area 3 are closer to the relevant base station 1 and therefore their communications shall be reasonably characterised by a lower "path loss" (they shall therefore be assigned a timeslot having lower priority Pi).

Now, making reference also to figure 2, we notice that an embodiment of the method according to the present invention includes an allocation algorithm, which is started for instance on the moment a mobile unit 2 requests a service Sx requiring the use of a given number Rx of channels Ci to a base station 1. The base station 1 measures the level and therefore the path loss PLx of the signal with which the mobile unit 2 has requested said service Sx on the receipt channel. On the basis of the path loss measured value PLx, the base station 1 attempts to allocate the Rx channels Ci in the timeslot having an increasing priority value Pi with the same attenuation PLx, in order that user equipment 2 transmitting signals having a high path loss use timeslots having a high priority value.

To this purpose, it is searched, starting from timeslots having higher priority values Pi, a timeslot Tx having Rx free channels Ci. If said timeslot Tx does not exist, the base station 1 refuses the requested service Sx to the mobile unit 2. If on the contrary said timeslot Tx is found, the base station 1 searches, if existing, a timeslot where at least a service employing Rx channels Ci is allocated among the timeslots with priority value Pi higher than that of the timeslot Tx. This search is made through a scanning based on a variable T cyclically decreased by one unit. If the variable T is zeroed, the requested service Sx is allocated in the timeslot Tx. If on the contrary a timeslot T is found where at least a service with Rx channels Ci is allocated a search is made among all the services employing Rx channels Ci and are in the same timeslot T, the service Sy showing the lower path loss PLm. At this point, the base station 1 compares the value of the lower path loss PLm found with that of the PLx path loss of the signal with which the mobile unit 2 has requested the service Sx to base station 1. If the PLx path loss value is lower than that of the PLm path loss, the requested service Sx is allocated in the timeslot Tx having Rx free channels Ci, otherwise it is allocated in the same the service Sy employing Rx channels Ci and showing the PLm path loss. In this last case, since a service having Rx channels Ci in the timeslot T got free, the algorithm described up to now is reiterated starting from this last timeslot, which is then identified as timeslot Tx in the reiteration of the algorithm itself, that is Tx = T.

When the service Sx is released, for instance after the interruption of a communication between mobile unit 2 and base station 1 or due to the transfer of a communication between two base stations 1 , it is possible to employ a release algorithm of essentially inverse type compared to the one described above to free the timeslots Ti with low priority values Pi.

Making reference to figure 3, we see that an embodiment of the method according to the present invention includes a release algorithm, which is started for instance on the moment at which a service Sx employing Rx channels Ci is released by a timeslot Tx. The base station 1 attempts therefore to allocate the Rx free channels Ci to a service Sy employing Rx channels Ci in the timeslot having the highest priority value Pi among those having lower priority value compared to that of the timeslot Tx. To this purpose, the base station 1 searches, if existing, a timeslot where at least a service employing Rx channels Ci is allocated, among the timeslots with priority value Pi lower than that of the timeslot Tx. This search is made through a scanning based on a variable T cyclically decreased by one unit. If said timeslot T is found, the service Sy characterised by the highest path loss amongst all the services employing Rx channels Ci in timeslot T is allocated in the timeslot Tx. In this last case, since a service having Rx channels Ci in the timeslot T has got free, the algorithm described up to now is reiterated starting from this last timeslot, which is then identified as timeslot Tx in the reiteration of the algorithm itself, that is Tx = T.

If the variable T is reset, the search can be terminated or, a service Sy employing a number of channels Ci lower than Rx newly searched among all the timeslots with a priority value Pi lower than that of the last timeslot Tx released. Said research is made through ad additional scanning based on a variable R cyclically decreased by one unit. Once this last variable is reset, the algorithm is terminated.

Other embodiments of the method according to the present invention can possibly include variants of said release algorithm, always started on the moment on which a service Sx employing Rx channels Ci is released by a timeslot Tx. For instance, instead of searching first the services Sy employing Rx channels Ci among all the timeslots having a priority value Pi lower than that of the timeslot Tx, to pass then to the search of services Sy employing a number of channels Ci lower than Rx always among all the same timeslots, it is possible to search the service Sy characterised by the maximum attenuation employing a number of channels Ci equal to or even lower than Rx in all the timeslots having lower priority value Pi compared to that of the timeslot Tx. With this algorithm, active services can therefore be reordered according to PLx attenuation values and increasing priority Pi values, irrespective of the number of channels Ci they employ.

At each allocation and/or release of a service, the priority values Pi assigned to the timeslots Ti can be recalculated. In place of the known algorithm based on the formula

Pi(k) = ^Pi(k- l) ₊ ^ , in another embodiment of the method according to the present invention it is possible to employ the following formula in which the past experience Pi(k-1) and the current situation s,(k) maintain a constant weight during the time:

Pi(k) = λPi(k - 1) + (1 - λ)s_i (k),

.where λ is a memory factor included between 0 and 1 , which can be freely selected according to the weight one wants to assign to the past experience or to the contingent situation. It is therefore clear that if λ tends to 0 or to 1 , the priority values Pi vary in a quicker or lower way, respectively, depending on the interference and/or quality measures of channels Ci by the base station 1.

A further development of this other embodiment can consist in calculating Sι(k) not on the basis of the simple statistics of the successful connections compared to total connections, but on the basis of the following formula:

N_^ - N _edι (k) in which N_{free i}(k) is the number of channels Ci which can be allocated with a good quality in timeslot i, N_maχ is the maximum number of channels (or codes) available per timeslot and N_used ι(k) is the number of channels currently already allocated in the timeslot i. Other embodiments and/or additions of the present invention may be made by those skilled in the art without departing from the scope thereof.

Claims

1. Method for the dynamic allocation of radio channels (Ci) in digital telecommunication networks with time division duplex access, in which radio signals are divided into frames having a pre-determined duration and each frame is divided into a pre-determined number of timeslots (Ti) which are assigned priority values (Pi) based on interference and/or quality measures of channels (Ci), each communication service (Sx) employing a particular number (Rx) of said channels (Ci) at a time, characterized in that it includes the following operational steps: - measuring the path loss (PLx) of the signal with which said communication service (Sx) has been requested;

- allocating said number (Rx) of channels (Ci) of the communication service (Sx) in a timeslot (Tx) having a priority value (Pi) increasing with the attenuation (PLx) of the relevant signal, in such a way that the services employing said number (Rx) of channels (Ci) are allocated in timeslots (Ti) having priority values (Pi) increasing with the attenuation of the relevant signal.

2. Method according to the previous claim, characterized in that at each request for a communication service (Sx) the services employing the same number (Rx) of channels (Ci) of the requested service (Sx) are reordered in such a way that the attenuation (PLx) increases with priority values (Pi).

3. Method according to one of the previous claims, characterized in that it includes an allocation algorithm including the following operational steps:

- searching, starting from timeslots (Ti) with highest priority values (Pi), a timeslot (Tx) having a number of free channels (Ci) equal to the number (Rx) of channels (Ci) of the requested service (Sx);

- searching, starting from timeslots with priority values (Pi) higher than that of the timeslot (Tx) found with this search, a communication service (Sy) having the same number (Rx) of allocated channels (Ci); - comparing the path loss values of the signals of the requested communication service (Sx) and of communication service (Sy) found with this search;

- allocating, according to the result of this comparison, one of these communication services (Sx, Sy) in the timeslot (Tx) having said number (Rx) of free channels (Ci).

4. Method according to the previous claim, characterized in that said algorithm is reiterated according to the result of said comparison between the attenuation values of the signals of the requested communication service (Sx) and of the communication service (Sy) found with this search.

5. Method according to claim 3 or 4, characterized in that it is searched, starting from timeslots with priority values (Pi) higher than that of the timeslot (Tx) found with this first search, the communication service (Sy) whose signals show the lower attenuation (PLm) among the communication services having the same number (Rx) of channels (Ci) allocated in the same timeslot (Tx).

6. Method according to one of the previous claims, characterized in that at each release of a communication service (Sx) are reordered according to increasing priority values (Pi) the services employing the same number (Rx) of channels (Ci) of the service released (Sx).

7. Method according to the previous claim, characterized in that it includes a release algorithm including the following operational steps:

- searching, among the timeslots with priority values (Pi) lower than that of the timeslot (Tx) of the released service (Sx), a timeslot (T) in which at least a communication service having the same number (Rx) of channels (Ci) of the communication service released (Sx) is allocated; - allocating in the timeslot (Tx) of the released communication service (Sx) the communication service (Sy) characterised by the highest attenuation among all the services employing Rx channels (Ci) in the timeslot (T) found with this research.

8. Method according to the previous claim, characterized in that it includes a release algorithm including the following additional operational steps:

- searching, among the timeslots with priority values (Pi) lower than that of the timeslot (Tx) of the released communication service (Sx), a timeslot (T) in which at least a communication service employing a number of channels (Ci) lower than that of the communication service released (Sx) is allocated; - allocating in the timeslot (Tx) of the released communication service (Sx) the communication service (Sy) characterised by a higher attenuation amongst all the services employing a number of channels (Ci) lower than that of the communication service released (Sx) and which are allocated in the timeslot (T) found with this search.

9. Method according to claim7 or 8, characterised in that said algorithm is reiterated starting from the timeslot of the last communication service released (Sy).

10. Method according to one of the previous claims, characterized in that at each allocation and/or release of a service, the priority values (Pi) assigned to the timeslots (Ti) are re-calculated on the basis of the following formula:

Pi(k) = λPi(k - 1) + (l - λ>, (*) , where k is the instant at which the service is allocated or released, Sj(k) is a logic function returning a number between 0 and 1 on the basis of the negative or positive result, respectively, of these requests for connection services and λ is a memory factor included between 0 and 1.

11. Method according to the previous claim, characterized in that S|(k) is defined by the following formula:

_Sj(k) = ^eβ'^{Λ } •}

in which N_frββ ι(k) is the number of channels Ci that can be allocated with a good quality in the timeslot i, N_max is the maximum number of channels (or codes) available for each timeslot and N_used ι(k) is the number of channels presently already allocated in timeslot i.

12. System for the dynamic allocation of radio channels (Ci) in digital telecommunication networks with time division duplex access, in which radio signals are divided in frames having pre-determined duration and each frame id divided into a pre-determined number of timeslots (Ti) which are assigned priority values (Pi) based on interference and/or quality measures of channels (Ci), each communication service (Sx) employing a particular number (Rx) of said channels (Ci) at a time, characterized in that at least one base station (1 ) for the reception and transmission of radio signals to a plurality of user equipment (2) includes means for the measurement of the path loss (PLx) of the signal with which said communication service (Sx) has been requested, as well as a control processor suitable to implement the method according to one of the previous claims.