RU2389139C2 - Information flow control in universal mobile telecommunication system (umts) - Google Patents

Abstract

FIELD: information technology.

SUBSTANCE: each data frame carried information belonging to one of several data streams. In the information flow control unit there is an estimation unit for determining, for each data stream at the end of first periods of time having a predefined first length, an estimation which is total number of received data frames, which were erroneous during the first period of time. The reference value calculation unit is connected to the estimation unit for calculating transmission capacity based on the estimated reference value, which determines the current maximum permissible transmission capacity for transmission from a data transmission unit to a data reception unit. The first unit can be a wireless network controller, and the second can be a base radio station. Data frames can be sent over a HS-DSCH through an Iub interface.

EFFECT: efficient algorithm for controlling information flow.

28 cl, 15 dwg

Description

FIELD OF THE INVENTION

The present invention relates to managing information flows for transmitting data frames from one node to another node, in particular to a network comprising a first node and a second node or data transmission node and a data receiving node, and to a method of transmitting data frames from the first node to the second node.

BACKGROUND

Universal Mobile Telecommunications System (UMTS) is a network technology that enables both voice and high-speed data transmission. It is part of the third generation (3G) wireless standards, as defined by the Third Generation Network Partnership Project (3GPP). Code Division Multiple Access (WCDMA), also called CDMA Broadband Access, is one wireless transmission method used in the UMTS system. UMTS is an evolution of the Global System for Mobile Communications (GSM) / General Packet Radio Service (GPRS), which supports packet voice and data.

The method, referred to as High Speed Downlink Packet Access (HSDPA) technology, is an extension of the UMTS system to increase the ability to transmit data, resulting in reduced cost per transmitted bit and greater spectral efficiency, in addition to significant increases in downlink data rates. HSDPA technology can provide an increase in current throughput by at least two or three times. It is based on the WCDMA standard and uses the same spectrum. HSDPA technology uses quadrature phase shift keying (QPSK) and 16 level quadrature amplitude modulation (16QAM).

The HSDPA technology method uses a distributed architecture to achieve low latency link adaptations by performing the most important process steps in radio base stations (RBS) and thus close to the wireless interface, see FIG. 1. HSDPA technology uses well-founded processing steps, including fast retransmission at the physical layer (L1) for erroneous packets, linking techniques, and link adaptation to provide improved packet data transmission.

The process steps of HSDPA technology mainly include:

- scheduling in base stations for the operation of packet data on the downlink;

- higher order modulation;

- adaptive modulation and coding;

- Hybrid automatic data retransmission requests (HARQ);

- feedback at the physical level about the current state of the channel;

- Transmission over a high speed downlink shared channel (HS-DSCH), allowing multiple users to share a wireless interface channel.

Some important features of HSDPA technology are described below.

1. Adaptive modulation and coding

HSDPA technology uses advanced link adaptation and adaptive modulation and coding.

2. Quick planning

In HSDPA technology, data flow is planned at base stations. HSDPA technology uses information about channel quality, terminal capabilities, quality of service (QoS) and power / code availability to achieve efficient scheduling of data packet transmissions.

3. Fast retransmissions at L1 level

When a communication error occurs, the mobile terminal immediately requests the retransmission of lost or erroneous data packets. This operation is designated as a method including hybrid automatic data retransmission requests (HARQ) to reduce delay and increase retransmission efficiency. HARQ request management is performed at the radio base stations, as illustrated in FIG. 2.

4. Channel quality feedback

In base stations in accordance with the method of HSDPA technology, channel quality estimates of each active user are collected and used. This feedback provides up-to-date information on a wide range of channel level physical state variables, including power control, acknowledgment ratio (ack) / negative acknowledgment (nack), quality of service (QoS), and user-specific HSDPA technology.

5. High Speed Downlink Shared Channels (HS-DSCH)

HSDPA technology is performed on high-speed downlink shared channels using a frame length of only two milliseconds compared to 10, 20, 40 or 80 ms frame lengths in previously used downlink shared channels (DSCH). Such shared downlink channels are downlink transport channels, each of which may be shared by several user equipments. A downlink shared channel is used to carry dedicated control or information data from a serving wireless network controller (SRNC). The DSCH will be paired with one or more dedicated downlink channels (DCH). HS-DSCHs provide 16-level quadrature amplitude modulation (16-QAM), link adaptation, and L1 retransmission aggregation using HARQ queries. HSDPA technology uses high-speed shared control channels (HS-SCCH) to carry the necessary modulation and retransmission information. High speed dedicated uplink dedicated physical control channels (HS-DPCCH) carry automatic data retransmission request (ARQ) confirmation messages, provide feedback on the quality of the downlink, and transmit other necessary control information on the uplink.

HSDPA technology requires an algorithm or method of controlling information flows that controls the transmission of data frames on the HS-DSCH, as defined, for example, in 3GPP Project TS 25.401, between the wireless network controller and the radio base station. The flow control algorithm is not standardized, but control messages, such as the “Capacity Allocation” message, are standardized. To control the flow of information, the RBS calculates the allocations to be transmitted in the “bandwidth allocation” messages sent to the wireless network controller (RNC), and the RNC sends data frames on the HS-DSCH to the RBS according to the information in the messages "bandwidth allocation", one bandwidth allocation for each data stream. When there is more data to send from the RNC, the User Buffer Size (UBS) information element (IE) in the HS-DSCH data frames is greater than zero. When the data frame frees the RNC controller buffer for the corresponding data stream, the UBS element is set to zero.

The flow control algorithm should control the constraints for both the wireless interface and the bandwidth of the HS-DSCH channel of the Iub interface, where the Iub interface is the interface between the RNC and the RBS.

In particular, the transmission of a data frame on the HS-DSCH from the RNC to the RBS is as follows. After the bandwidth (capacity) is provided to the RNC through the RBS, as obtained from the bandwidth allocation control frame or from the initial bandwidth allocation control frame received from the RBS, as described in 3GPP TS 25.433, and the RNC there is data waiting to be sent, a data frame is used to transmit data on the HS-DSCH. If bandwidth is provided to the RNC by the RBS using the initial bandwidth allocation control frame, as described in 3GPP TS 25.433, this bandwidth is only valid for transmitting the first data frame on the HS-DSCH. When data is awaiting transmission and a bandwidth allocation control frame has been received, the data frame on the HS-DSCH will be immediately transmitted in accordance with the received allocation, that is, using the bandwidth corresponding to this allocation. Each data frame sent on the HS-DSCH includes a user buffer size information element for indicating the amount of data waiting for a corresponding stream for a designated priority level.

When carrying out efficient transmission of data frames from the RNC to the RBS, the Iub interface bandwidth limitations must be considered. When dedicated channels are connected, for example, for voice, the bandwidth for the HS-DSCH channels will be less, provided that, as will be assumed in the future, these dedicated channels and the HSDPA stream share the same physical channel. If the bandwidth for HS-DSCHs causes the loss of data frames of the HS-DSCH, this has a bad effect on TCP based applications. Then it is better to lower the effective bit rate for HS-DSCH to a level where the data frame loss ratio becomes acceptable.

SUMMARY OF THE INVENTION

The objective of the invention is to provide an effective algorithm for managing the flow of information.

Another object of the invention is to provide an information flow control algorithm that adapts a data frame stream so that data frames can be sent without too high a loss, and so that the buffers in the receive node are always validly filled.

The flow control (information) algorithm in the RBS station to control HS-DSCH channel traffic through the Iub interface is used to ensure that the RBS station priority queues always have the right amount of data. If it is assumed that there is data in the RNC to send to the RBS, the priority queues should neither become empty nor overflow. Too long buffers give too long delays, too short buffers give empty buffers when the user unexpectedly schedules. The information flow control algorithm is aimed at managing priority queue flows in such a way that they become stable.

There are two bottlenecks: the wireless interface and the Iub interface. Both of these bottlenecks are taken into account in the information flow control algorithm. Available high-speed bandwidth through the Iub interface varies greatly. If too much high-speed traffic is distributed through the Iub interface, frame loss and long delays degrade the performance of high-speed packet data.

The RBS flow control algorithm for controlling HS-DSCH channel traffic through the Iub interface uses a reference value to determine the available bandwidth for HS-DSCH channel traffic; this reference value is dynamically adapted by taking into account detected traffic errors.

The advantage of providing a dynamic reference value for high-speed throughput on the Iub interface is that large HS-DSCH data frame loss can be avoided.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention is described below by way of non-limiting embodiments with reference to the accompanying drawings, in which:

Figure 1 is a schematic diagram of communication between a radio base station and user equipment using high speed downlink data transmission,

2 is a schematic diagram illustrating differences in data transmission to user equipment using a downlink shared channel in accordance with two different methods,

Figure 3 is a diagram illustrating the transfer of data from an application server to an application running in user equipment using HSDPA technology and the flow control function of information to control the Iub interface,

4 is a diagram illustrating various levels in communication between an RNC and an RBS using HSDPA technology,

5 is a transmission diagram illustrating the principle of managing information flows using standardized messages,

6 is a schematic diagram illustrating buffers in an RNC, an RBS, and a user equipment UE,

Figa is a diagram illustrating the calculation of the state of the thread of the priority queue,

Fig. 7b is a flowchart of the steps performed in the calculation shown in Fig. 7a,

Figa - state machine changes the state of the thread of the priority queue,

Fig. 8b is a flowchart showing the steps of the change process shown in Fig. 8a,

Fig.9 is a diagram of the reduction ratio as a function of time for one traffic error,

Figure 10 is a diagram similar to Figure 9 for two traffic errors,

11a is a state machine illustrating the calculation of a variable representing the maximum combined bit rate,

11b is a flowchart of the steps performed in the calculation shown in FIG. 11a,

12 is a schematic illustration of a radio base station illustrating an information flow control unit and corresponding memory cells.

DETAILED DESCRIPTION

With reference to FIGS. 3-6, a general information flow is described in a system constructed in accordance with a universal mobile communication system (UMTS), including both in a mobile telephone network and in some other network, for information transmitted from another user equipment (UE) networks in a mobile telephone network using high-speed downlink packet access (HSDPA) technology, in particular a stream between a wireless network controller (RNC) and a radio base station (RBS) in a mobile telephone network with ides. These drawings mainly include only blocks and functions that relate to or are necessary for the said general flow of information.

Many packet data applications use the standardized Transmission Control Protocol (TCP) for data transfer defined by the Internet Standards Working Group (IETF). As shown in FIG. 3, data, for example, an Internet page, can be sent from the application server 101 via a public data network (PDN) or a common network 103, such as the Internet, to an application 105 running on user equipment 107, then is in the mobile terminal. The transmission control protocol of the working group according to the standards for the Internet has its own window size, which limits the number of bytes in various buffers that data must pass between the application server 101 and the user equipment. The wireless link control sublayer (RLC) has a different window size. Automatic retransmission request (ARQ) confirmation messages are used both in accordance with the TCP protocol and at the RLC sublayer to control the correctness of the transmissions.

The public network 103 is connected via a wireless access network (RAN) (not shown) to the RNC controller 111 in the gateway or support node 109, for a GPRS system this node includes a GPRS network support gateway (GGSN) and a serving GPRS network support node (SGSN). A GPRS network support gateway is a router that serves as a gateway or interface between packet data networks and mobile telephone networks, in particular between a packet data network such as Internet Protocol (IP) network (103) and a serving network support node GPRS on a mobile phone network 104.

The interface between the core (core) network and the wireless access network (RAN) is usually called the Iu interface, and the packet interface between the SGSN and the RNC 111 is called the Iu-PS interface. The interface between the RNC and the RBS station 113 is called the Iub interface, and the interface between the RBS station 113 and the mobile terminal 107 is called the Uu interface.

Packets received from the gateway or support node 109 are first stored in the SDU buffers 115.

The information flow control (FC) function is used to control the communication between the RNC controller 111 and the RBS station 113, in particular the data frame stream on the HS-DSCH using the Iub interface, and aims to keep 127 priority (PQ) queues in the RBS station short and Do not overfill the Iub interface transport network, that is, the transport network between the RNC and the RBS, see network parts 119, 121 in FIG. 3.

Iub Interface Architecture for HSDPA

The flow control function includes parts 123, 125 located in the RNC controller 111 and in the RBS station 113, respectively. In the RBS station, it is part of the function (function block) 126 medium access control for HSDPA (MAC-hs). It interacts using the Iub interface protocol messages transmitted in the Iub interface control frames, with the information flow control part 123 in the RNC, which is part of the dedicated media access control (MAC-d) function (function block) 124 in the controller 111 RNC, see also Figure 4.

The streams controlled by the information flow control function 123, 125 are MAC-d protocol data unit (PDU) streams transmitted in HS-DSCH data frames in accordance with the Iub interface frame protocol (FP).

Each MAC-d PDU arriving at part 125 of the information flow control function in the RBS station 113 is stored in one of the priority queues 127 to wait for selection via the scheduler function (function block) 129 of the RBS station scheduler for transmission over the Uu interface to the user equipment 107.

At RBS 113, one priority queue 127 is provided for each MAC-hs connection of the HS-DSCH of the connected user equipment 107, and one managed stream of MAC-d PDUs is provided through the Iub interface for each priority queue. Each such stream is located in a flow control function designated as a priority queue stream (PQF). The priority queue stream is defined as packets arriving for the same user having the same contents of the Common Channel Priority Indicator (CmCH-PI) field, as defined in the standard documents. In almost all cases, for each user equipment there is only one priority queue stream, which then is the downlink information stream for the corresponding user, although in general there can be many priority queues for each connected user equipment 107.

Each priority queue thread is transported over the Iub using one instance of the frame protocol (FP) of the Iub using a dedicated ATM adaptation layer 2 (AAL2) connection as a unidirectional transport channel.

Figure 4 illustrates the configuration of the level and the corresponding elements of the interfaces Iub and Uu.

- Part 401 of the wireless link control sublayer of the RNC 111 is primarily intended to provide a lossless communication line, that is, a reliable communication line over a wireless interface for transmitting data based on the TCP protocol. It provides reliability using error detection and retransmission. The RLC sublayer 410 interacts with the wireless link control sublayer portion 402 included in the RLC / MAC-d portion 133 of the user equipment 107, see FIG. 3.

- The MAC-d function 124 in the RNC controller 111 interacts with the MAC-d function 403 included in the RLC / MAC-d part 133 of the user equipment 107, see FIG. 3. In addition, the MAC-d function of the RNC controller 111 is illustrated here as including an HS-DSCH frame protocol (FP) processing unit 404. This frame protocol processing unit, in turn, includes the information flow control part 123 and interacts with the information flow control part 125 included in the frame protocol (FP) processing unit 405 of the HS-DSCH channel of the MAC 126 function 126 of the RBS station 113. The RBS station's MAC-hs function includes a flow control part 125, a scheduler 129, and a MAC-hs HARQ function 131, see FIG. 3. The MAC-hs function 126 in the RBS interacts with the MAC-hs function (function block) 135 in the user equipment 107. The user equipment MAC-hs function 135 includes the HARQ function (function block) 136, see also FIG. 4 feature 131 MAC-hs HARQ stations 113 RBS.

- AAL2 / ATM VC layer having parts 407, 409 in the RNC 111 and in the 113 RBS station.

- Parts 411 of the physical layer (L1) in the controller 111 RNC, parts 413, 415 in the station 113 RBS and part 417 in the user equipment 107.

For communication (exchange) between the RBS station 113 and the RNC controller 111, a transport network 419, such as an ATM network and / or a PDH / SDH network, is used as compared to the transport network portions 119, 121 of FIG. 3. A wireless network 421 is used to communicate between the RBS station 113 and the user equipment 107.

Flow control messages between the RNC and the RBS

HSDPA data, that is, MAC-d PDUs, is sent from the RNC controller 111 to the RBS station 113. Each MAC-d stream of a given priority level is equal to one priority queue stream and is represented by one queue 117 in the RNC and one priority queue 127 in the RBS station. Several MAC-d PDUs belonging to one MAC-d stream are sent in each frame protocol data frame (FP) of the HS-DSCH.

Data frames sent over the Iub interface for each priority queue stream are a stream controlled using bandwidth allocation (CA) messages sent in control frames from the RBS station 113 to the RNC controller 111, see FIG. 5. The bandwidth allocation message determines the bit rate as the number of MAC-d PDUs that are allowed to be transmitted during a predetermined time period for the priority queue flow in question.

For a simple case, the RBS station 113, based on the fill level of the corresponding buffer, i.e. the length of the corresponding priority queue in the RBS station, on the basis of the state of the wireless interface, i.e. the state related to the Uu interface, and based on the congestion of the transport network in the Uu interface, decides the bit rate that must be allocated for a given priority queue for use by the RNC controller 111 for transmission on the corresponding HS-DSCH. The RNC controller generates data streams in accordance with the latest received bandwidth allocation messages. Message structures can be found in 3GPP TS 25,435, in particular in FIG. 21A “Data Frame”, in FIG. 36 “Bandwidth Allocation” and in the accompanying text.

The information flow control function block 123, 125 is aware of the average data rate available for the priority queue stream over the wireless interface between the RNC controller 111 and the RBS station, or at least an estimate of said average data rate. He also knows the number of PDUs from this priority queue stream that are waiting in the RBS buffer for this queue 127. Based on this information, the information flow control functional block may decide to change the distributed speed of the priority queue in question. The main goal is to keep the target number of pending PDUs in the RBS 113, that is, not too many and not too few PDUs in each priority queue.

The buffers for queues 117, 127 are designed such that PDUs are most likely lost only in the transport network of the Iu interface or in the wireless Uu interface.

Purpose of using the information flow control function

The Iub interface information streams in HS-DSCHs are a stream controlled by the MAC-hs function 126 information flow control function 125 in the RBS station 113. Iub interface protocol messages that can be used to control information flows are defined in 3GPP TS 25.435 (Iub interface). The information flow control function itself is not standardized.

The purpose of the information flow control function (function block) is to maintain an “acceptable” number of MAC-d PDUs buffered in RBS 113, that is, to keep RBS station priority queues 127 short enough for retransmissions of the RLC sublayer, but long enough to ensure planned throughput.

The same RLC sublayer logical buffer for priority queue flows can be considered distributed between the RNC controller 111, the RBS station 113, and the user equipment 107. The MAC-d PDUs that need to be retransmitted have a higher priority in the RNC than PDUs to be sent for the first time from the RNC, see FIG. 6. Therefore, part of the RBS of the RLC sublayer, priority queues 127, must be "short" or not too long, this is one reason for using the information flow control function to control HS-DSCH transmission from the RNC 111 to the RBS 113.

HS-DSCH channel traffic is transmitted as the best possible quality of service (QoS) in the transport network 119, 121; part 415 between the RNC controller and the RBS station. The information flow control function should adjust the HS-DSCH channel traffic flow so that the loss of MAC-d PDUs due to too long Iub interface transport delays, for example, caused by congestion in the transport network, is acceptable. There is a tradeoff between high frame loss together with high throughput and low frame loss with lower bandwidth.

There are mainly two bandwidth bottlenecks for HSDPA traffic in transport networks between the RNC controller 111, the RBS station 113, and the user equipment 107, which should both be considered in the flow control function:

- Iub interface,

- Uu interface, i.e. a radio interface, or a wireless interface.

HS-DSCH data frames, each of which carries one or more MAC-d PDUs, are transmitted through AAL2 in AAL2 packets. These packets are transmitted using ATM technology "sliding way". Thus, each ATM cell has an "initial byte" belonging to the AAL2 architecture, which determines where the first AAL2 packet begins in the ATM cell. This is used when losing an ATM cell in order to find where the next proper AAL2 packet begins. ATM cells are transmitted over the transport network 119, 121; a portion 419 between the RNC controller 111 and the RBS station 113. If the traffic flow exceeds the available throughput of the AAL2 path, the transport network, you may have to discard some AAL2 packets or ATM cells due to buffer overflows in the transport network. When one or more AAL2 packets (level) of the data frame are lost, the data frame reassembled at the RBS will have an erroneous cyclic redundancy check (CRC) checksum for the payload and an erroneous length that can be detected by the RBS. Losses of frames can also be caused by bit errors due to poor transmission quality, which also lead to CRC errors, but in most such cases the frame length will be correct, and this makes it possible to distinguish the most probable causes of erroneously received data frames.

The following describes a frame loss algorithm for detecting an Iub interface congestion condition by observing cases of incomplete or defective HS-DSCH channel data frames arriving at RBS 113, the algorithm generates a flag indicating that the loss of at least one data frame was detected during a predetermined time period, for example, a 100 ms interval, for each priority queue thread. When the number of data frames lost for the priority queue stream exceeds a predetermined loss threshold for a predetermined period, the estimated bitrate allocation rate caCalcBitrate for this priority queue stream temporarily decreases by a predetermined fraction, for example, by 50% of its nominal value, see FIG. 9 and 10, and is sent to the RNC controller 111 in a bandwidth allocation (CA) message.

When RBS 113 detects too many data frame losses in all PQFs for another longer period of time, such as one second, the value of targetHsRate decreases. The value of this variable indicates the estimated total bit rate available for transmitting data frames from the RNC 111. This variable provides a reference value that indicates the maximum combined bit rate value that can be used for HSDPA technology HS-DSCH data frames via the downlink Iub interface.

When the iubFrameLossFlag flag is detected, showing at least one frame loss for the priority queue stream for a predetermined period of time, the input data is the Iub interface HS-DSCH data frames from the priority queue stream, see Fig. 7a. Each data frame is evaluated and a flag is set if the data frame is defective or detects an error.

Detection of lost and / or erroneous data frames is performed by checking each received data frame, as shown in the following pseudo-code fragment:

If the header CRC and payload CRC in the data frame are correct, accept the data frame and do not set a flag.

If the CRC of the header in the data frame is incorrect, discard the frame and set the iubFrameLossFlag flag.

If the header CRC is correct and the payload CRC in the data frame is incorrect, discard the data frame and check if the length of the data frame, including the payload CRC, in bytes is equal to the value given by the formula:

INT [(4+ (MAC-d PDU Block Length) +7) / 8] * "NumOfPDU" +9.

If true, an incorrect CRC of useful data was most likely caused by poor transmission quality and no further action is taken.

If this is false, the incorrect payload CRC was most likely caused by the loss of one or more AAL2 packets, and the iubFrameLossFlag flag is set.

Therefore, detecting one or more erroneous data frames over a period of 100 ms sets the iubFrameLossFlag flag. It should be noted that it is impossible to determine how many destroyed (damaged) frames were detected over a period of time.

The steps of the detection procedure are illustrated in FIG. 7b. Thus, when the data frame is received, the procedure begins and in the first step 703 they ask if the CRC checksum of the header field of the data frame is correct. If this is true, at step 705 they ask if the CRC checksum of the payload field of the data frame is correct. If this is true, at step 707, a data frame is received. If it was determined at step 705 that the checksum CRC of the payload field is not correct, at step 709 it is determined whether the length of the data frame is correct, that is, whether it is equal to the standard length as defined by the formula above. If this is true, at block 711, a data frame is discarded. If it was determined in step that the data frame length is not correct, the iubFrameLossFlag flag is set in step 713, and then step 711 is performed. If it was determined in step 703 that the CRC checksum field of the data frame header field is not correct, steps 713 are also performed and 711.

After at least one indication of Iub interface congestion has been detected for a period of 100 ms, as indicated by the setting of the iubFrameLossFlag flag, the value of the variable caCalcBitrate, which indicates the bandwidth allocation for the priority queue stream, is temporarily reduced by 50% , and then there is a reverse increase to 100% of the level during the subsequent overload-free predefined period of time, for example one second. Bandwidth allocation is related to the targetHsRate variable, which will be described below.

This reduction procedure 801 is illustrated in FIG. 8a and is performed at the end of each predetermined time period. It uses the value of the iubFrameLossFlag flag as input and generates a reduction coefficient, or iubCoeff coefficient, which is specific to the PQF stream under consideration, as well as a modified value of the iubHsTrafficErrors variable, which gives an estimate of the number of data frame errors, as determined for the RBS station in question, which occurred from the beginning of the current longer period of time having a length of, for example, 1000 ms. Thus, the iubCoeff coefficient is generally a function of the frequency of occurrence of errors of high-speed traffic of the Iub interface. The bandwidth allocation (CA) bit rate for a PQF stream decreases when a high-speed Iub interface traffic error occurs for the stream.

When the iubFrameLossFlag flag is set, that is, it is TRUE, the value of the variable caCalcBitrate is reduced using the iubCoeff coefficient. The coefficient iubCoeff is initially and usually equal to the coefficient maxCoeff, which has a value of 100%. When a high-speed Iub interface traffic error occurs during a predefined shorter period, the iubCoeff coefficient is always set to minCoeff, for example, equal to 50%, in the next shorter period of time. The value of the iubCoeff coefficient changes between the values of minCoeff and maxCoeff so that it is never lower than 50% and never higher than 100%.

When no errors of high-speed Iub interface traffic for the PQF flow occur for a shorter time, the value of the iubCoeff variable is increased by a predefined CoeffStep step, for example, equal to 5% relative to 100%.

In the example illustrated by the diagram in FIG. 9, after one high-speed Iub interface traffic error that is not followed by an error, the iubCoeff coefficient is set to its minimum value of 50% immediately after the high-speed Iub interface traffic error indicated by the iubFrameLossFlag flag set. This is followed by an increase in the iubCoeff coefficient by 5% in each interval with a duration of 100 ms. After 1 second, the iubCoeff coefficient value returned to its maximum value of 100%. Lower iubCoeff coefficients produce correspondingly reduced bandwidth allocation bit rates.

In another example, illustrated by the diagram in FIG. 10, two Iub interface traffic errors occurred for the PQF stream for another shorter time period of 100 ms, each time reducing the bandwidth allocation bit rates due to decreasing the iubCoeff coefficient to level of 50%.

The procedure steps for calculating the iubCoeff coefficient for each PQF stream and simultaneously changing the value of the error counter iubHsTrafficErrors are illustrated in Fig. 8b. Thus, at the end of each shorter period of time and for each thread, the priority queues are first checked at step 803 whether the iubFrameLossFlag flag has been set. If it was set, in step 805, the IubCoeff coefficient for this priority queue flow is set to its minimum value minCoeff. Then, the iubHsTrafficErrors counter is incremented by one at step 807, and finally, at step 809, the iubFrameLossFlag flag is reset, that is, it is set to FALSE. If it was determined in step 803 that the flag was not set, in step 811 it is determined whether the coefficient iubCoeff is less than its maximum value maxCoeff. If this is true, the coefficient is increased by the CoeffStep value in the step. After the steps have been completed, the bandwidth allocation for the priority queue flow is reduced in some way using the current value of the iubCoeff coefficient, as indicated by step 815.

The iubHsTrafficErrors variable is specific to the RBS station in question. This is a counter representing all the Iub interface high-speed traffic errors that occurred for all PQF flows in all shorter periods of time over a longer period of time.

One mistake of high-speed Iub interface traffic here is equivalent to detecting that one iubFrameLossFlag flag is set for one PQF stream for a shorter period of time.

In the example, each PQF thread adds a maximum of 10 to the iubHsTrafficErrors variable in one second. In general, all PQF flows in a 113 RBS station can, for example, add hundreds of increments per second, for example, if there are many high-speed Iub interface traffic errors.

Thus, every time the iubCoeff variable for the PQF stream is set to its minimum value, even if it has already been set to its lowest value, the iubHsTrafficErrors variable is incremented by one. The value of the iubHsTrafficErrors variable is reset after reading it, see the description with reference to Figs. 11a and 11b. This is done at the end of each long time interval by means of step 1101 of calculating the target speed.

The targetHsRate variable represents the maximum combined rate of caCalcBitrate, which can be distributed to all PQF streams in the same RBS station. The value of targetHsRate is limited by the minimum value of minHsRate and the maximum value of maxHsPvate and has an initial value of maxHsRate. The calculation of targetHsRate is illustrated in FIGS. 11a and 11b.

The targetHsRate variable is sensitive to too many Iub interface high-speed traffic errors. If, for example, 5 or more errors of high-speed Iub interface traffic were detected within one second in any of the PQF flows in the RBS station, the value of the targetHsRate variable is reduced by the decrement value of targetHsStepDown, for example, equal to 2% relative to the maximum value of maxHsRate. If less than 5 errors per second were detected over a longer period of time, for example 1 second, the value of the targetHsRate variable is increased by the value of the increment targetHsStepUp, for example, equal to 1% relative to the maximum value of maxHsRate.

The current value of the targetHsRate variable is used as the reference value when calculating the bit rate of the bandwidth allocation. The bandwidth allocation for the priority queue, for example, can be calculated as

CA = (targetHsRate) · (iubCoeff) / (sum of all iubCoeff).

Procedure 1101 for calculating the value of the targetHsRate variable has the input parameters minHsRate, maxHsRate, hsRatelubErrorThreshold, hsRateRecoverTime, targetHsStepUp and targetHsStepDown. It has an iubHsTrafficErrors variable at the input and produces a new value for the targetHsRate variable at the output. The targetHsRate output variable is basically a function of the IubHsTrafficErrors variable, and is assigned a value at the end of each longer period of time by executing the procedure, as shown in the following snippet of pseudocode:

IF iubHsTrafficErrors> = hsRatelubErrorThreshold

TO

- decrease targetHsRate by targetHsStepDown if more than minHsRate

- set errorFreeSecCounter = 0

ELSE

increase errorFreeSecCounter by 1

IF errorFreeSecCounter> = hsRateRecoverTime

TO

- increase targetHsRate by targetHsStepUp if less than maxHsRate

- set errorFreeSecCounter = 0

ELSE

End IF

assign iubHsTrafficErrors = 0

The steps of procedure 1101 are also illustrated by the diagram in FIG. 11b. In a first step 1103, it is determined whether the number of traffic errors specified by the current value of the iubHsTrafficErrors counter is greater than some small acceptable threshold value called hsRatelubErrorThreshold. If not more, then another counter, called errorFreeSecCounter, to count the number of longer consecutive time periods in which the number of errors was valid and thus did not exceed the threshold value, increase by one at step 1105. Then, at step 1107, check more whether the current value of this counter is other than a threshold value called hsRateRecoverTime. If this is true, it means that there has been an acceptable number of traffic errors for a sufficiently long time, and step 1109 is performed to determine if the current value of targetHsRate is equal to its maximum value maxHsRate. If this is not true, step 1111 is performed, in which the current value of targetHsRate is increased by targetHsStepUp. Then, at 1113, the errorFreeSecCounter counter for longer time periods having a valid number of traffic errors is reset. Finally, the ubHsTrafficErrors counter for the total number of errors is reset at step 1115. If it was determined at step 1109 that the targetHsRate was equal to maxHsRate, then steps 1113 and 1115 were also performed. If at step 1107 it was determined that the errorFreeSecCounter was greater than hsRateRecoverTime, then also performed 1115. If it was determined in step 1103 that the current value of iubHsTrafficErrors is greater than hsRatelubErrorThreshold, step 1117 is performed to check if the current value of targetHsRate is equal to its minimum value of minHsRate. If it is larger, the current value is reduced by the value of the targetHsStepDown parameter in step 1119. Then, steps 1113 and 1115 are performed as described above. If it was determined in step 1117 that the targetHsRate value is minHsRate, then steps 1113 and 1115 are performed.

To perform the information flow control function, the information flow control unit 1201 and special memory cells must be added to the RBS station 133, as shown in the diagram in FIG. 12. The memory cells include, for each priority queue 127, and therefore, for each priority queue thread, a memory cell 1203 for storing the iubFrameLossFlag flag value and a memory cell 1205 for storing the iubCoeff reduction factor. The memory cells 1207, 1209 and 1211 are provided for fixed values of the parameters maxCoeff, minCoeff and CoeffStep used in calculating the coefficients. Memory cell 1213 contains the current value of the iubsHsTrafficErrors counter, and another cell 1215 contains the current value of the targetHsRate variable. Memory cells 1217-1227 contain fixed parameter values used in calculating the value of the targetHsRate variable. Memory cell 1229 contains the current value of the errorFreeSecCounter counter for error-free periods. The information flow control block 1201 comprises a block 701 for detecting possible frame loss, a block 801 for calculating the iubCoeffs coefficients and increasing the error counter iubTrafficErrors, a block 1101 for calculating or changing the targetHsRate variable, and a block 1231 for calculating the bandwidth allocation based on the coefficients and the value of the targetHsRate variable.

Claims (28)

1. A mobile communication network including a data transmission node and a data receiving node, wherein data frames are transmitted from the data transmission node to the data receiving node, each data frame carries information belonging to one of the plurality of data streams, containing in the receiving node
an estimation determination unit for determining, for each of the data streams at the end of the first time periods having a predetermined first length, an estimate representing the total number of data frames received from the transmission node that were erroneous during the first time period, and
a reference value calculation unit connected to the evaluation determination unit for calculating, based on the determined estimates for all data streams, a bandwidth reference value that defines the currently permissible total maximum bandwidth for transmission in all data streams from the data transmission node to the data receiving node. 2. The network according to claim 1, characterized in that the reference value calculation unit is configured to reduce the bandwidth reference value by a predetermined decreasing value if the determined estimate is greater than or equal to the threshold value, in the case when the bandwidth reference value is greater than the minimum value . 3. The network according to claim 1, characterized in that the reference value calculation unit is configured to increase the bandwidth reference value by a predetermined increasing value if the determined estimate is less than the threshold value, in the case when the bandwidth reference value is less than the maximum value. 4. The network according to claim 1, containing in the receiving node a detection unit for determining upon receipt of each data frame whether the received data frame is erroneous. 5. The network according to claim 1, characterized in that the data transmitting node is a wireless network controller and the data receiving node is a base station in a universal mobile communication system (UMTS). 6. A mobile communication network including a data transmission node and a data receiving node, wherein data frames are transmitted from the data transmission node to the data receiving node, wherein each data frame carries information belonging to one of the plurality of data streams, the network includes receiving node
a unit for determining a possible loss of frames for determining, for each of the data streams at the end of the second time periods having a predetermined second length, whether any of the data frames received during the second time period is erroneous, and
a bandwidth allocation unit connected to a possible frame loss determination unit for distributing for each of the data streams for which said determination unit determined that the data frame was erroneous, a reduced predetermined fraction of the total available bit rate or bandwidth for transmission from the node data transmission to the data receiving node,
characterized in that the bandwidth allocation unit is configured to, in the case where the reduced predetermined portion for the data stream has been allocated in the immediately preceding second time period, increase the reduced predetermined proportion at the end of each subsequent second time period by a predetermined step to a maximum predetermined a certain fraction in the case when no data frame was erroneous for the corresponding period of time. 7. The network of claim 6, comprising a counter, said bandwidth allocation unit being configured to increase the counter when a reduced predetermined fraction has been allocated to the data stream, the counter value being an estimate representing the total number of received data frames, which were erroneous. 8. The network according to claim 7, containing a reference value calculating unit for calculating a bandwidth reference value defining a maximum currently allowed bandwidth for transmission from a data transmission node to a data receiving node, at the end of the first time periods having a predetermined first length greater than the second length, based on the value of the counter. 9. The network of claim 6, comprising a detection unit for determining upon receipt of each data frame whether the received data frame is erroneous. 10. The network according to claim 6, characterized in that the data transmission node is a wireless network controller and the data reception node is a radio base station in a universal mobile communication system (UMTS). 11. A data receiving unit for a mobile communication network including a data transmitting unit, the data transmitting unit transmitting data frames to the data receiving unit, each data frame carries information belonging to one of the plurality of data streams, comprising
an estimation determination unit for determining, for each of the data streams at the end of the first time periods having a predetermined first length, an estimate representing the total number of received data frames that were erroneous during the first time period, and
a reference value calculation unit connected to the evaluation determination unit for calculating, based on the determined estimates for all data streams, a bandwidth reference value that defines the currently permissible total maximum bandwidth for transmission in all data streams from the data transmission node to the data receiving node. 12. The data receiving unit according to claim 11, characterized in that the reference value calculating unit is configured to reduce the bandwidth reference value by a predetermined decreasing value if the determined estimate is greater than or equal to the threshold value, in the case when the bandwidth reference value is larger minimum value. 13. The data receiving unit according to claim 11, characterized in that the reference value calculating unit is configured to increase the bandwidth reference value by a predetermined increasing value, if the determined estimate is less than the threshold value, in the case when the bandwidth reference value is less than the maximum value . 14. The data receiving unit according to claim 11, comprising a detection unit for determining upon receipt of each data frame whether the received data frame is erroneous. 15. The data receiving unit according to claim 11, characterized in that the data transmitting unit is a wireless network controller and the data receiving unit is a radio base station, the data transmitting unit transmits data frames via HS-DSCH channels in a universal mobile communication system (UMTS). 16. A data receiving unit for a mobile communication network including a data transmitting unit, the data transmitting unit transmitting data frames to the data receiving unit, wherein each data frame carries information belonging to one of the plurality of data streams, the data receiving unit includes
a possible frame loss determination unit for determining for each of the data streams at the end of the second time periods having a predetermined second length whether any of the data frames received during the second time period is erroneous, and
a bandwidth allocation unit connected to a possible frame loss determination unit for distributing for each of the data streams for which the determination element determined that the data frame was erroneous, a reduced predetermined fraction of the total available bit rate or bandwidth for transmission from the transmission node data to the data receiving node,
characterized in that the bandwidth allocation unit is configured to, in the case where the reduced predetermined portion for the data stream has been allocated in the immediately preceding second time period, increase the reduced predetermined proportion at the end of each subsequent second time period by a predetermined step to a maximum predetermined a certain fraction in the case when no data frame was erroneous for the corresponding period of time. 17. The data receiving unit according to claim 16, comprising a counter, said bandwidth allocation unit being configured to increase the counter when a reduced predetermined fraction has been allocated to the data stream, the counter value is an estimate representing the total number of received data frames that were erroneous. 18. The data receiving unit according to claim 17, comprising a reference value calculating unit for calculating a bandwidth reference value defining a maximum currently allowed throughput for transmission from the data transmission unit to the data receiving unit at the end of the first time periods having a predetermined a first length greater than a second length based on a counter value. 19. The data receiving unit according to clause 16, comprising a detection unit for determining upon receipt of each data frame whether the received data frame is erroneous. 20. The data receiving unit according to claim 16, wherein the data transmitting unit is a wireless network controller and the data receiving unit is a radio base station, the data transmitting unit transmits data frames via HS-DSCH channels in a universal mobile communication system (UMTS). 21. A method of transmitting data frames from a data transmission node to a data receiving node in a mobile communication network, each data frame carries information belonging to one of the plurality of data streams, comprising the steps of:
determining for each of the data streams at the end of the first time periods having a predetermined first length, an estimate representing the total number of received data frames that were erroneous during the first time period, and
on the basis of certain estimates for all data streams, a reference value of the throughput is determined, which determines the currently permissible total maximum throughput for transmission in all data streams from the data transmission node to the data receiving node. 22. The method according to item 21, wherein in the calculation step, in the case where the bandwidth reference value is greater than the minimum value, the bandwidth reference value is reduced by a predetermined lowering value if the determined estimate is greater than or equal to the threshold value. 23. The method according to item 21, wherein in the calculation step, in the case when the reference value of the throughput is less than the maximum value, the reference value of the throughput is increased by a predetermined increasing value if the determined estimate is less than the threshold value. 24. The method according to item 21, characterized in that at the stage of determination determines when receiving each data frame, whether the received data frame is erroneous. 25. A method of transmitting data frames from a data transmission node to a data receiving node in a mobile communication network, wherein each data frame carries information belonging to one of the plurality of data streams, the method comprises the steps of:
determining for each of the data streams at the end of the second time periods having a predetermined second length whether any of the data frames received during the second time period is erroneous, and
allocate for each of the data streams for which the determination result is that the data frame was erroneous, a reduced predetermined fraction of the total available bit rate or throughput for transmission from the data transmission node to the data receiving node,
characterized in that after the distribution step, in the case where a reduced predetermined fraction has been distributed for the data stream, this reduced predetermined fraction is increased at the end of each subsequent second period of time by a predetermined step up to the maximum predetermined fraction in the case when none the data frame was not erroneous during the corresponding time period. 26. The method according A.25, characterized in that at the distribution stage, in the case when a reduced predetermined proportion has been distributed for the data stream, the counter is increased, and the counter value is an estimate representing the total number of received data frames that were erroneous. 27. The method according to p. 26, containing the step of calculating the reference value of the bandwidth that determines the maximum currently allowed bandwidth for transmission from the data transmission node to the data receiving node, at the end of the first time periods having a predetermined first length greater than the second length, based on the value of the counter. 28. The method according A.25, characterized in that at the stage of determination determines, upon receipt of each data frame, whether the received data frame is erroneous.

Images