WO2008150206A1

WO2008150206A1 - Dual random access channels in extended range

Info

Publication number: WO2008150206A1
Application number: PCT/SE2007/050403
Authority: WO
Inventors: David Astely; Tobias Tynderfeldt; Per Skillermark; Erik Westerberg
Original assignee: Telefonaktiebolaget Lm Ericsson (Publ)
Priority date: 2007-06-07
Filing date: 2007-06-07
Publication date: 2008-12-11

Abstract

The present invention relates to cellular communication systems and to the random access procedure. An uplink random access channel RACH is available in each cell for out-of-sync terminals to be able to access the system. A RACH window must be long enough to accommodate bursts send sent from terminals at the cell boarder as well as from terminals close to a receiving base station. In general the RA-window is dimensioned for a basic cell range. The RACH capacity is further dimensioned by the frequency allocation and the period of repetition of the RA-window. A problem is if the geographical coverage is to be extended to cover more than the basic cell range, because the length of the RA-window needed for terminals at the cell range will take much of the uplink transmission capacity, to the disadvantage of other traffic. The present invention solves the problem with two RACH1 with different window lengths. One RACH intended for terminals in the inner cell and the other for terminals in the outer cell. The advantage is the capacity of each RACH can be dimensioned in proportion to the load of RACH bursts.

Description

DUAL RANDOM ACCESS CHANNELS IN EXTENDED RANGE

TECHNICAL FIELD OF THE INVENTION

The present invention relates to cellular radio communications systems, and to the random access procedure and channel. In particular it relates to a method for a radio base station, and a method for a mobile terminal for efficient use of the resources allocated for terminals to randomly access a cellular network. The invention also relates to a radio base station and to a mobile terminal that are adapted for performing the respective methods.

DESCRIPTION OF RELATED ART

For cellular systems adapting a time frame structure on uplink channels, a radio base station must receive signals from a terminal within a scheduled frame, time slot, or time window. Therefore a terminal must adjust its transmission timing to compensate for the propagation delay to the radio base station. The base station that supports a link with the terminal sends time alignment commands for the terminal to adjust its transmissions timing. The terminal is then synchronised to the network.

If, for a period, there is no transmission from the terminal the base station cannot time align the terminal and it gets out of synchronisation. This happens if there has been no transmission from the terminal within about 500ms. How long the terminal is synchronized depends both on the stability of its internal clock and its mobility and resulting change in radio path to the RBS. The inactivity periods when the terminal goes out-of-synch occur quite frequently in packet switched communication systems, due to its inherent irregular traffic patterns and time periods of low traffic. Typically the terminal is also out-of-sync when a user wants to set up a voice call if there is no other link established at the call initiation.

An out-of-sync terminal accesses the network via a random access procedure. Initially the radio base station finds a strong broadcast channel and detects information on it relating to where a random access channel is allocated. The allocation is typically a recurrent time window on a specified frequency carrier. The terminal then sends an RA-burst (Random Access burst) in the RA-window (Random Access window). A problem is though, that there is a propagation delay for the broadcast signal to arrive at the terminal and vice versa for the signal sent from the terminal to arrive at the base station. The longer the distance between the terminal and base station, the longer the propagation delay. This result in that a random access sent by a terminal close to the base station will arrive at the base station earlier than a random access sent from a terminal distant to the base station. Figure 1a is a view of a radio base station 10, a distant terminal 11 , and a close terminal 12. Figure 1b is a diagram with a horizontal axis indicating distance from the base station 10 to the distant and close terminals 11 , 12, and a vertical axis indicating time. In the diagram an arrow A_11-0L indicates the propagation of a DL (Down Link) signal from the base station 10 to the distant terminal 11 , an other arrow A_11-UL indicates the propagation for the RA-burst sent from the distant terminal 11 , to the base station 10, i.e. in the UL (Up Link) direction. The corresponding delays t11_DL and t11_DL+UL are marked on the time axis. In the diagram there is also corresponding arrows A₁₂_DL_, A_{12 UL}, and time delays t12_DL, t12_DL+UL marked on the time axis for the close terminal 12.

Owing to the non-compensated propagation delays t11_DL+UL, t12_DL+UL the RA-window need be long enough to accommodate access bursts a terminal 11 at the cell border and close terminal 12 and the respective propagation delays. Terminals at a distance corresponding to a propagation delay longer than the RA-window can not access the system.

When dimensioning a RACH (Random Access CHannel), in addition to the RA-window length also the allocation of frequency band, and

RA-window repetition interval must be considered. These parameters dimension the capacity of the RACH₁ in the number of RA-bust possible to transmit. The capacity need be balanced to the amount of resources occupied that could alternatively be used for other communication than RA-bursts.

The problem and solution of the present invention applies to all types of cellular systems that have a frame structure on the uplink radio channels. However, for the further description of the problem, example will be given with respect to an LTE system. It should be understood that the LTE is one example of several possible and used for increased understanding of the background of the invention.

The LTE is standardized by 3GPP and is an abbreviation of 3G Long Term Evolution. In LTE there is an FDD mode and a TDD mode. In the TDD mode the same frequency carrier is used for both UL and DL transmission, by alternating time window allocated for respectively the UL and DL. The alternating UL and DL time windows T_DL T_UL are illustrated in a time line of figure 2. It represents the use of a frequency carrier. The multiple radio access technology used on the LTE UL is Single Carrier FDM. Different users and control channels are scheduled OFDM symbols in the frequency domain, i.e. occupying specific sub-carriers of the total frequency carrier, and are scheduled in the time domain for specific number of OFDM symbols.

An operator of an LTE system would typically dimension the RACH capacity to the number of access attempts per second. The capacity of the random access channel is typically set by the bandwidth in time and frequency dimension dedicated to random access communication. A too low capacity of the RACH would degrade the access performance due to collisions with access bursts from different users, while a too large RACH capacity wastes UL transmission resources.

Let us now consider a standard-size cell in a reasonably high-traffic area. With existing state-of-the art the operator would set the RA- window length to match the cell size, in LTE to 0.5 ms. The operator would then dimension the RACH capacity with respect to the relatively high access burst intensity induced by the high traffic in the area. Let us assume that this leads to two RA-windows on one 1.25 MHz frequency band repeated every 5 ms.

For the standard cell size this works fine, but let us now consider what happens if this operator wants to greatly extend the size of the cell, up to 150 km, for the purpose of coverage into an area with very little traffic. With the existing solution the operator would need to extend each RA-window to 1 ms. This would double the amount of UL resources consumed by the RACH. Since the number of access bursts would increase only slightly due to the vast majority of the traffic is generated in the inner region of the cell, while only little traffic is generated in the outer region, the RACH is now over-dimensioned by 100%. Figure 3 illustrates an UL time period T_UL, with a basic RA- window and an extended RA-window overlapping the basic RA- window. RA-bursts are illustrated as dots within the RA-windows, the vast majority of them are received within the basic RA-window, while in the extended RA-window a single RA-burst is received.

The problem is thus that extending the cell size for the purpose of coverage becomes very costly in terms of RACH capacity allocation. It should be noted that the cost is not at all balanced to the amount of traffic in extended coverage. This means that just re-configuring the cell to take larger coverage reduces the UL capacity of the cell even in the absence of any traffic in the extended area.

SUMMARY OF THE INVENTION

The present invention solves the above problem by a method for a base station serving a cell, and the method provides a first and a second RACH in the cell, with the first RACH having a RA-window shorter than that of the second RACH. The base station further broadcasts information on the two RACH to be used by terminals for selection of an appropriate of the RACH. The present invention also comprises a base station adapted for performing the method.

The first RACH is intended for terminals in the radio base station vicinity and the second RACH is intended for terminals distant from the base station. Then invention further relates to a method for a terminal on the selection of RACH for sending random access bursts. The method comprises the steps of the terminal detecting broadcast information on RACH selection and selecting a RACH that matches an expected distance to the base station. The invention further comprises a terminal adapted for performing the method.

An advantage with the present invention is the cell coverage can be extended without the RACH matching the cell size will have to occupy a large amount of the transmission resources. The operator may still provide capacity for the first RACH in relation to the traffic demand, for example by means of the repetition interval of the RA-window or the amount of frequency spectrum provided. This decreases the initial cost if the operator wants to expand the coverage outside locations were most of the traffic is generated.

DESCRIPTION OF THE DRAWINGS

Figure 1a is a view of a cell, radio base station and two terminals.

Figure 1b is coordinate system, indicating propagation time on vertical axis as function of time on horizontal axis, for arrows representing signal transmissions to and from two terminals.

Figure 2 is a time line, indicating time window on a frequency carrier for UL and DL transmissions.

Figure 3 is a UL time window, part of it allocated for a basic and a extended RA-window.

Figure 4 is a view of a cell, with a basic range R1 , and an extended range R2, a base station and a terminal. Figure 5a - 5c are diagrams in the time/frequency domain of scheduled UL and DL periods and allocation of RA-windows.

Figure 6 is a flowchart of the steps performed by a mobile terminal.

DESCRIPTION OF PREFERRED EMBODIMENTS

The core of the present invention is to provide two RACH (Random Access CHannels) in cells with extended range. A first of the two RACH has a RA-window length adapted for receiving RA-burst from terminals in the vicinity of a base station 10 serving the cell. Figure 4 is a view of a radio base station 10 with a basic cell range R1 indicated by a dotted circle an extended cell range R2 the maximum extended range R2 indicated by a continuous drawn circle. Figure 4 also discloses a terminal 12. Typically the maximum distance of a terminal to access the first RACH is the basic cell range R1. The second RACH has a longer RA-window than the first RACH and can receive RA-bursts from terminals in the extended range beyond the basic cell range R1.

Figure 5 a-c discloses three examples of RACH allocation in a TDD mode LTE system. In the examples, access to the frequency carrier is alternating given to UL and DL time windows intersected by guard periods. The first example, figure 5a, discloses an example of how a RACH is scheduled according to the prior art in a normal range cell. In the UL time windows T_ULl RA-windows 41 of a first length are scheduled on two sub-frequencies of the frequency carrier, and recurring in each UL time window T_UL. The two RA-windows have a length of 0.5ms and will be repeated every 5ms. In a second example, figure 5b, the cell has been extended, and by applying the prior art the O

RA-windows 41 have been extended to the double length, i.e. 1.0 ms, to accommodate also RA-burst from the extended cell range. A drawback is the RA-windows 42 occupy a large amount of the UL transmission capacity.

Figure 5 c, illustrates scheduling of the inventive two RACH in the cell. A first RACH is scheduled with two RA-windows 41 of 0.5 ms repeated every 5ms. This is just as in the case with basic cell range because it is expected the RA-bursts generated within the basic cell range will remain also after the cell expansion. In addition also a second RACH is scheduled in one expanded RA-window 42 of 1.0 ms and that is repeated in each second UL time window T_UL, i.e. every 10 ms. In the example the second RA-window 42 overlap with the first RA-window 41.

By separating the first and second RACH the operator can dimension the first RACH to match the load of RA-bursts from the basic cell range, by means of the number of RA-windows and their repetition interval. The risk of RA-bursts from two terminals colliding is decreased, and if it happens there is only a short interval until the next opportunity to repeat the RA-burst. Thereby the majority of terminals in the vicinity of the base station 10 can access the network within short when they attempt to. The terminals in the outer range have longer periods until they can repeat their RA-burst in a next RA-window 42. Due to they are few the risk of their RA-bursts colliding is low.

Having defined a first and a second random access channels with different duration in the time domain the second part of the invention relates to the selection of one of the two RACH when a terminal shall send a RA-burst (Random Access burst). The objective is the first RACH shall be selected for transmission of the vast majority of the random access bursts from terminals close to the RBS. The reason is first random access channel has the highest capacity and shortest access time. The second RACH shall be used for the RA-bursts from terminals in the extended part of the cell because though it has lower capacity it has a sufficient time window to receive random access bursts with high propagation delay. The complication in the selection of the RACH is to make some type of estimation of the distance between the terminal 12 and the base station 10.

In its basic form, it is expected that the terminal 12 is within the basic range R1 of the cell, and only if the first RA-bursts fail from being acknowledged, the second RACH may be used for sending further RA-bursts from the terminal 12. In the basic embodiment the base station 10 broadcasts information on the two RACH including at least the relative lengths of the respective RA-windows 41, 42, the frequency and time domain allocation of the two RACH. Figure 6 illustrates in detail the steps performed by the terminal 12 in the basic embodiment of selecting one of the two RACH for transmission of a RA-burst. In the first step 601 , the terminal 12 detects the broadcast information on the RACH. This is made when the terminal 12 enters a new cell or is powered on. In the next step, 602, the terminal enters active mode. The active mode may have been entered before the terminal entered the cell, and thus the order of steps 601 and 602 may be swapped. The terminal enters state A, see step 603, and that means the first RACH is to be used for any transmission of RA-burst. Next, 604, the need to transmit a RA-burst is checked. If, there is no such need the terminals remains in state A, in a loop back to step 603. If, however, there is a need to transmit a RA-burst, a counter N is set o zero, 605, and a first RA-burst is transmitted, 606 on the first RACH. The counter N is increased by one, 607, and receipt of an acknowledgement from the base station 10 is checked 608. If an acknowledgement is received, the terminal remains in state A, step 603, and the further steps followed in loop. Absent any acknowledge in step 608, the counter N is checked 609, and on condition the counter N does not exceed 3, the step 606 re-entered and a further RA-burst transmitted. The steps following on step 606 are entered in a loop. If, however, the result of check in step 609 is more than 3 RA- burst have been sent without any acknowledgement have been received, i.e. N > 3, state B is entered, 610. In state B the terminal only uses the second RACH with the long RA-window 42 for sending RA-bursts. The next RA-burst is sent, 611 , on the second RACH, the counter N increased, 612, by one, and any receipt of acknowledge from the base station is checked 613. If no acknowledge is received the transmission, 611 , of a RA-burst on the second RACH is repeated until a check in step 614 detects more than 6 RA-bursts have been sent. Then no further RA-burst may be sent until a period T2 has expired, 619, and then the method step 603 is re-entered in a loop. If₁ in step 613 an acknowledge is detected, the terminal remains, 615, in state B, the counter N is set, 616, to zero, and the need to send a RA- burst is checked, 617. If, there is such a need, the step 611 , is re- entered and the RA-burst is sent over the second RACH, the steps following in a loop. If, no need for sending a RA-burst is detected in step 617, the step of 615, remaining in state B is re-entered in a loop. Tree optional steps, 618, are indicated by checked boxes 620, 621 , 623 in figure 6. After, a last sent RA-burst has been acknowledged in step 613, a timer T1 is set 620. The timer is checked, 621 , and if no further RA-burst has been sent until the timer expires, the terminal re- enters state A in step 603, and the loop continues with the steps following on 603.

It may happen that the method disclosed in figure 6 is interrupted because the terminal 12, exits the active mode. The optimal timer T1 may still be active, and if it has not expired when the terminal 12, re- enters, 602, the active mode, the terminal enters state B₁ when the second RACH is used. This is made by an optional step, 623, checking if the optional timer T1 has expired.

In figure 6, the counters N are set to 3 and 6 for maximum number of RA-bursts to be transmitted on first the first RACH and then the second RACH. These numbers are just examples and that can be chosen from a wide range.

The terminal 12 located within the basic range R1 of the cell, and acting according to the method in figure 6 will have an access procedure that is essentially unaffected by the extension of the cell. In contrast, if the terminal 12 is located in the outer part, beyond the basic range R1 while within the extended range R2, it will fail the first 3

RA-burst send over the first RACH when being in state A. However, when entering state B the RA-burst will be successfully received by the base station 10. The price paid by the terminal 12 when being in the outer part of the cell is a slightly longer first-access time, from maybe 10 ms to a multiple of 10 ms depending on the configuration.

Note that this is only the first access when the terminal is not in synch. Any sub-sequent access, made before the timer T1 has expired, will not be prolonged because of use of the first RACH when the second

RACH should have been used. Accordingly, in order to let for example a VoIP call have short access time the timer T1 shall be long enough to accommodate for voice and varying transmission delay interruptions.

The method described in connection with figure 6, may alternatively be further developed to increase the estimation accuracy of the distance to the base station 10. In an alternative, the time alignment adjustment commands received from the base station 10 are integrated by the terminal and used for estimating the distance and selection of one of the two RACH. Terminals estimated to be within the basic cell range, albeit when taking the estimation inaccuracy into account could be outside the basic range, should select the second RACH if short access time is preferred.

During a period of not being in active mode, and getting out of sync, a timer may be used for determining whether the estimated distance is reliable or not. The timer value may also be adjusted in relation to how close to the basic cell range boarder the terminal is located according to the estimation. Thus if the terminal is within but close to the basic cell range R1 , the estimated distance should only be used within a relative short time span.

Alternatively to the scheduling in figure 5c where the second RA- window,42, overlap in time with some occurrences of the first RA- window, 41 , the second RA-window 42 can be selected to be separate from the basic RA-window 41. The first part of the second RA-window, 42, may then be used for receiving RA-bursts from terminals within the basic range of the cell. A terminal within the basic range is then free to select the RA-window, 41 , 42 that first occur in time irrespective of it being within the first or the second RACH. The terminal is constructed for selecting one of the two RACH according to any of the alternative solutions described above. Alternatively, the terminal is programmed for adapting to a selected of the alternative methods. The base station 10, in the alternative, in addition to broadcasting information on the two RACH, with the cell system information, also broadcast rules for how the terminal shall select one RACH. A set of rules is predefined and enables the terminal to adapt its selection of RACH in accordance with the broadcast rules. An operator may then apply a different set of rules in different cells. The rules are selected for situations when the terminal is expected to be in either the vicinity of the base station 10 or distant. The operator can then adapt the capacity of the first and second RACH in relation to the expected traffic generated in the inner and in the outer region of the cell, and steer the terminals right to access the respective first and second RACH in relation to the expected generated traffic and capacity .

The above embodiments are given in an LTE system, however, the invention may be implemented also in other cellular systems for example GSM or WCDMA. In GSM a physical channel consist of one time slot of 8 available in a repetitive TDMA frame. In GSM an extended cell range is provided in the prior art by allocating two consecutive TDMA time slots for RACH. A complication is the RACH is a logical channel and plural logical channels are mapped on one physical channel. The prior art extended GSM cell RACH consumes a huge part of the LJL transmission resources, corresponding to the situation in figure 5b. If the present invention is applied to the case of an extended range GSM cell, the extended RA-window should be scheduled as well as a second RA-window. In contrary to the prior art GSM extended range RA-window 42, the extended RA-window 42 can be scheduled to be repeated with long intervals, while the short RA- window 41 is repeated more frequent.

Also in WCDMA logical channels are mapped on physical channels by a frame structure. In WCDMA the signal is spread over the total frequency carrier. Transmission resources is varied not only by the length of the RA-time windows, their repetition period but also with the transmission power. These parameters can be used in differentiating the transmission resources allocated to the respective first and second RACH, while still ensuring the extend RA-window 42 is long enough to accommodate RA-bursts transmitted from the extended cell outer range.

Claims

1. A method for a radio base station (10) comprising the steps of: -providing a first random access channel, by allocating a recurrent first time window (41 ) with a first length; characterised by the further steps of:

-providing a second random access channel, in a recurrent second time window (42) with a second length, longer than the first length; -broadcasting information, relating to at least the occurrence in time of respectively the first and second time window (41 , 42) and their relative window lengths.

2. The method according to claim 1 , wherein the first random access channel is provided for higher capacity in the number of random access bursts possible to be received.

3. The method according to claim 2, wherein said higher capacity is provided by means of the first time window occurrence is more frequent than the second time window occurrence.

4. The method according to claim 2 or 3, wherein said higher capacity is provided by means of the first random access channel is allocated more radio frequency spectrum than the second random access channel.

5. The method according to claim 1 , wherein said information comprises rules relating to situations when a specific of the first and second random access channel shall be selected by a terminal.

6. The method according to claim 5, wherein the rule is that the terminal shall send one or more random access burst/s on the first random access channel if there is no indication on the second random access channel is to be preferred over the first random access channel.

7. The method according to claim 6, wherein the rule is that the terminal shall send one or more random access bursts over the second random access channel if no acknowledge has been received by the terminal on a predetermined number of random access burst sent by the terminal.

8. The method according to claim 5, 6 or 7, wherein the rule is that the terminal shall send one or more random access burst/s on the one of the first and second random access channel that was last used by the terminal for transmitting a random access burst and that was acknowledged by the base station.

9. The method according to claim 8, wherein the rule further includes the terminal shall send random access burst on the second random access channel only if the time lapsed since the last acknowledged random access burst was transmitted is shorter that a predefined time period.

10. The method according to any of the preceding claims wherein the recurrent second time window overlap in time and frequency with at least some of the recurrences of the first time window.

11. A radio base station arranged for providing, - a first random access channel with a first window (41) length; characterised by being arranged for further supporting: - a second random access channel with a second window (42) length larger than the first; and arranged for, - broadcasting information on the first and second random access channel, relating to at least their relative window lengths and respective occurrence in time.

12. A method for a mobile terminal (12), comprising the steps of,

- detecting (601) broadcast information relating to a first random access channel, characterised by the further steps of:

- detecting information on a second random access channel in the broadcast information, wherein the information on the first and second random access channel relates to at least time occurrences and relative lengths of respective reception windows;

-selecting (609, 614, 621 , 623)one of the two random access channel in relation to an expected distance to a radio base station broadcasting the information;

-transmitting (606,611 ) a random access burst on the selected of the first and second random access channels.

13. The method according to claim 12, wherein the terminal initially selects the first random access channel.

14. The method according to claim 12, wherein the terminal initially selects either of the first and second random access channel.

15. The method of claim 14 wherein according to the detected information the reception window (42) of the second random access channel is not overlapping in time with the reception window (41) of the first random access channel.

16. The method of claim 12, 13 or 14 wherein the terminal selects (609) the second random access channel if a predefined number of random access bursts have been transmitted on the first random access channel without any acknowledge being received.

17. The method of claim 12, 13, 14, or 16 wherein the one of the first and second random access channel is selected (613,615) on which a last acknowledged random access burst was transmitted.

18. The method of claim 17 wherein the second random access channel is selected only if the time lapsed (621 , 623) since the last acknowledged random access burst was transmitted is shorter than a predefined time period.

19. The method of claim 12 wherein the expected distance to the base station is estimated by integrating received time alignment commands, and selecting the second random access channel when the estimated distance is in the region of or larger than a basic cell range.

20. A mobile terminal (12) adapted for performing the method of claim 12 or as claimed in any claim dependent on claim 12.