WO2002073901A1 - Method and device for transmission of packet data traffic on a multiplexed link - Google Patents

Method and device for transmission of packet data traffic on a multiplexed link Download PDF

Info

Publication number
WO2002073901A1
WO2002073901A1 PCT/SE2001/000512 SE0100512W WO02073901A1 WO 2002073901 A1 WO2002073901 A1 WO 2002073901A1 SE 0100512 W SE0100512 W SE 0100512W WO 02073901 A1 WO02073901 A1 WO 02073901A1
Authority
WO
WIPO (PCT)
Prior art keywords
packets
delay
packet
streams
transmission
Prior art date
Application number
PCT/SE2001/000512
Other languages
French (fr)
Inventor
Szabolcs Malomsoky
Original Assignee
Telefonaktiebolaget Lm Ericsson (Publ)
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Telefonaktiebolaget Lm Ericsson (Publ) filed Critical Telefonaktiebolaget Lm Ericsson (Publ)
Priority to PCT/SE2001/000512 priority Critical patent/WO2002073901A1/en
Publication of WO2002073901A1 publication Critical patent/WO2002073901A1/en

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic control in data switching networks
    • H04L47/10Flow control; Congestion control
    • H04L47/22Traffic shaping
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic control in data switching networks
    • H04L47/10Flow control; Congestion control
    • H04L47/32Flow control; Congestion control by discarding or delaying data units, e.g. packets or frames
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W28/00Network traffic management; Network resource management
    • H04W28/02Traffic management, e.g. flow control or congestion control
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W28/00Network traffic management; Network resource management
    • H04W28/02Traffic management, e.g. flow control or congestion control
    • H04W28/0247Traffic management, e.g. flow control or congestion control based on conditions of the access network or the infrastructure network
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W72/00Local resource management
    • H04W72/50Allocation or scheduling criteria for wireless resources
    • H04W72/54Allocation or scheduling criteria for wireless resources based on quality criteria

Definitions

  • the present invention relates to transmission of information packets in a communication system for example a radio access network and especially to a method for transmission of packet streams and to a node handling such streams.
  • UTRAN is a short name of the UMTS (Universal Mobile Telecommunication System) Radio Access Network.
  • a mobile terminal MT, a radio base station BS and a radio network controller RNC are three basic nodes in UTRAN, and shown in figure 1.
  • the mobile terminal MT is connected by a radio link to the radio base station BS .
  • From the radio base station BS radio links with several mobile terminals MT are provided.
  • the radio network controller RNC is connected to several radio base stations BS by fixed links, however in figure 1 just one exemplar of each of the nodes is disclosed.
  • Communication in the direction to the mobile terminal MT i.e. in the downlink direction, is set up via the radio network controller RNC and via the radio base station BS to the mobile terminal MT .
  • the communication is also conveyed from the communication source via an exchange connected to the radio network controller RNC, however, no nodes except the radio network controller RNC, the radio base station BS and the mobile terminal MT will be described here. Communications in the opposite direction, i.e. in the uplink direction, is transmitted via the same nodes.
  • Figure 2 shows protocol stacks up to layer 2/3 of the radio network controller RNC, the radio base station BS and the mobile terminal MT .
  • the protocol stack shows layers of protocols that correspond to specific functionalities for establishing communication.
  • FIG. 1 Also shown in figure 1 are the fixed link interface Iub between the radio network controller RNC and the radio base station BS, and the air interface Uu between the radio base station BS and the mobile terminal MT.
  • the lower layers of the protocol stacks are related to physical and functional requirements of transmitting data from the sending to the receiving node.
  • the lowest layer on both sides of the air interface Uu is thus WCDMA (Wideband Code Division Multiple Access) that corresponds to the radio multiple access technology.
  • WCDMA Wideband Code Division Multiple Access
  • the transport network infrastructure LI is depicted, e.g. an ATM/AAL2 network or an IP network. Any of these is used for transferring MAC frames.
  • the MAC frames are produced by the above MAC (Media Access Control) layer.
  • the MAC layer as well as the RLC (Radio Link Control) layer are located in the radio network controller RNC and in the mobile terminal MT.
  • the radio links shall be utilised optimally because the capacity of these radio links is limited.
  • Streams of information packets in the downlink direction are formatted into streams of radio frames with a certain length and format in the radio network controller RNC.
  • the main function of the MAC layer in the downlink is to schedule the radio frames.
  • the scheduling of the radio frames is done periodically in every 10, 20, 40 or 80 ms to adapt to the radio interface.
  • QoS Quality of Service
  • Various connections 33 are multiplexed in a queuing system, which consists of a FIFO (First In First Out) buffer 32 of size b and a server 31 with rate S.
  • the queuing delay packets suffer when waiting in the buffer must be below a certain limit. As long as the packets of the received flows are evenly distributed over time and the total flow rate are not higher than the serving rate of the server 31, all packets will be served the moment they arrive. No queuing delays will thus arise. If, however, the traffic flow is bursty, i.e.
  • packets are unevenly distributed over time, while the packet flow rate is the same on a time averaged basis, packets will have to be queued in the buffer 32 before being served.
  • the more uneven distribution of the packets the longer the buffer 32 need to be. Packets in the buffer 32 will be delayed, the longer queue of packets the longer delay. Packets delayed longer than maximum are deleted and thereby lost.
  • Leaky-bucket like shaper is not applicable in UTRAN because multimedia traffic, and especially voice traffic can not be a subject of such shaping because its real-time character requires low delays.
  • Leaky-bucket like shaping is usually applied to non-real time traffic to modify the packet arrival patterns such that it conforms to a simple, mathematically tractable traffic description.
  • Figure 4a shows the multiplexing of packet streams on the transport network LI of fixed interface Iub. In this example there are four voice connections, each corresponding to a packet stream on one of the horizontal lines.
  • the multiplexing also corresponds to the order in which the packets will arrive at the fixed interface Iub.
  • the packets are scheduled periodically with a constant interarrival time TTI (Transmission Time Interval) .
  • TTIs are indicated by vertical dotted lines.
  • the TTI are 20 ms long for voice, however, different services have different TTIs.
  • the example with voice service on the Iub link can - be generalised to all types of services and also to other links .
  • phase offset 42 is selected for the connection.
  • the phase offset 42 is selected randomly, according to a uniform distribution, over the TTI.
  • the objective of the phase offset 42 selection is to smooth the traffic and thereby decrease burstiness.
  • the packet transmission is started after the phase offset 42 is selected.
  • the packet transmission can be modelled with a periodic on-off source. Packets are transmitted during the "on-periods" and not transmitted during the "off-periods" . In figures 3, 4a and 5, small boxes indicate packet transmissions in the on-periods.
  • Figure 5 shows four connections 53 with periodic packet arrivals multiplexed in a queuing system, which consists of a server 51 and a FIFO buffer 53.
  • the buffer temporarily stores the arrived packets if the server 53 is busy when they arrive.
  • the buffer depth b corresponds to the maximum delay on the fixed interface Iub as required by the QoS. If the buffer is full when a packet arrives the packet will be lost .
  • the present invention addresses a problem that packets are lost when periodic scheduling is used on a transmission link.
  • the present invention solves the problem by a method introducing a random delay to each packet in the transmitting node.
  • the delay is randomly selected within an interval and causes the burstiness to decrease and and the arrival order of packets of different streams to vary. Thereby the maximum server queue is reduced and, moreover, the delay will 'be more fairly distributed over the various packet streams.
  • a node for an electronic information system comprising an input for receiving information streams and a packet output link and that is arranged to transmit the information streams as packet streams scheduled periodically on the packet link.
  • the transmitter is further arranged to introduce a small randomly selected delay to each packet, and thereby causing the packets to jitter on the link.
  • the invention is also solved by a shaper arranged to introduce a s random delay to periodically scheduled packets that are to be transmitted on a multiplexed link as streams of packet data.
  • the invention has the great advantage of reducing the packet losses without having to increase the receiving server capacity.
  • a further advantage is that losses are distributed more fairly over the streams.
  • a further advantage is that the invention is easy to implement with respect to the protocol stack used in communication systems such as UTRAN.
  • the functions to introduce the random delays will then be implemented between those handled by the MAC-layer and those handled by the Transport Network Layer.
  • Figure 1 is a block diagram of basic nodes in a radio access system.
  • Figure 2 is a block diagram of UTRAN protocol layers.
  • Figure 3 is a block diagram showing packet streams fed to an input buffer of a receiving server.
  • Figure 4a is a diagram of scheduling of 4 packet streams on a link according to prior art.
  • Figure 4b is the same as figure 4a but with the invention applied.
  • Figure 5 is a diagram corresponding to that of figure 3 but with the packets scheduled periodically and time divided into transmission intervals.
  • Figure 6 is a flow chart showing the introduction of a random delay on the packets.
  • Figure 7 is a time diagram showing phase relations of the packets of 4 different streams.
  • Figure 8 is a a block diagram of a packet data traffic shaper.
  • Figure 9 is a diagram showing packet losses as a function of the maximum length of the random delay introduced.
  • Figure 10 is a diagram showing the gain with respect to non introduced delay as a function of maximum length of the random delay introduced.
  • FIG. 7 shows the prior art phase relations of 4 packets of different streams that are scheduled and multiplexed on the same link.
  • Each packet stream corresponds to a connection, i.e. a voice connection or a multimedia connection.
  • a packet P1-P4 from each of the stream is transmitted within the same TTI (Transmission Time Interval) .
  • TTI Transmission Time Interval
  • the server 51 has capacity for handling just one packet P1-P4 at the time.
  • the third and the fourth packets P3, P4 will be buffered before being served and the fourth packet P4 will be buffered the longest time.
  • the phase relations remain the same for subsequent TTIs, all with respect to prior art.
  • FIG. 4b is the same as figure 4a except for that the transmission moment for the packets P4 are slightly varying between different TTIs as a result of the random delay introduced to the packets P.
  • the random delay interval 41 is also indicated in figure 4b.
  • FIG. 6 is a flowchart showing the steps performed by different protocol layers of a transmitting node, in the UMTS system.
  • a first step 61 the packets of various streams are scheduled periodically and deterministic according to the principle disclosed with reference to figure 4a. In short this means that there is an opportunity for each stream to have a packet scheduled at a precise occasion within each TTI. The precise moment is determined by the phase offset 42 of each packet stream. The scheduling is handled by the MAC-layer of the protocol stack shown in figure 2.
  • a small random delay is introduced to each packet P4.
  • the delay is randomly selected from an interval 41 that is much smaller than the TTI.
  • the packets P4 are multiplexed on the link. This is handled by the Transport Network layer, i.e. layer 1 of the protocol stack shown in figure 2.
  • the small delay introduced in step 62 causes packets to jitter on the link.
  • jitter is meant that the arrival times differs slightly and randomly from the expected arrival times.
  • the jitter will be in the region of [phase offset; phase offset + X] where X denotes the maximum delay.
  • Figure 8 discloses the shaper 81 that comprises a random number generator 82, a table 83, a clock
  • MAC frames are encapsulated into Iub frames and sent towards the transport layer, they are first entering the random-delay shaper 81.
  • the shaper 81 attaches a temporary identifier to the MAC frames and timestamps the arriving MAC frames.
  • the random number generator 82 generates random numbers according to a uniform distribution from the interval 41 [0ms ; Xms] , X being the maximum introduced delay.
  • a random delay is generated for each MAC frame and the random delay value is added to the timestamp of the MAC frame.
  • the temporary frame identifiers and the related timestamps (updated with the random delay) are then inserted into the table 83, which is ordered according to the timestamp values.
  • the MAC frames are stored in the buffer 84. If the internal clock 85 of the shaper (81) reaches the value of the first timestamp in the table (83), the MAC frame with the corresponding temporary identifier is sent to the transport network.
  • the receiving server buffer 51 depth need to be reduced correspondingly to the maximum random selected delay introduced. This could suggest that the advantages with the introduced random delay would not compensate for the disadvantage of the corresponding reduction of the receiving buffer depth 51.
  • compressed voice packets are scheduled from the MAC layer every 20 ms (TTI) , they can be delayed at most 5 ms on the Iub interface, and the average packet loss should be at most around 0.1% - 1%.
  • 387 voice connections i.e. packet streams, are multiplexed over a 2720 kbit/sec link.
  • the voice connections have random phase offsets 42.
  • Voice packets are sent periodically in the 20ms TTIs according to an on-off process.
  • the length of the on period is exponentially distributed with a mean of 10 sec.
  • the probability of the on period is 0.5.
  • the codec is a 12.2 kbps codec, it sends 31 bytes in the on periods. In the off periods, 8 bytes are sent in every 8 th TTI.
  • the parameters of the voice model are not very important in our evaluation. The most important is that the voice and other services in UTRAN can be modelled as an on-off source with deterministic arrivals.
  • the voice packets are multiplexed in an AAL2 multiplexer and transported in ATM cells to the Base Station.
  • the loss performance results are depicted, for various maximum lengths of the random delay applied to each packet.
  • the horizontal axis of the diagram corresponds to the buffer 51 size that was decreased from 5ms to 1ms, while the uniform distribution range controlling the randomisation was increased from [0ms; 0ms] to [0ms; 4ms] .
  • the random delay introduced in the Random Delay Shaper step 72 was chosen according to a uniform distribution over the range [0ms; 2ms] . This way, the maximum 5ms-delay requirement was met in all cases.
  • the uniform black line shows the average loss and the dashed line shows the maximum loss for the various buffer depths.
  • the uniform black line shows the gain related to the average packet loss (AVGGAIN) while the dashed line shows the gain regarding maximum packet loss (MAXGAIN) , compared to applying a buffer 51 depth of 5 ms and no random delay.
  • the horizontal axis corresponds both to the buffer size and to the maximum random delay, in the respect that the maximum total delay of both shall be 5 ms .

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Quality & Reliability (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The present invention relate to packet data transmission between nodes in a radio access system. The packets are scheduled periodically and smoothed over time by applying a corresponding phase offset (42) for each connection. A problem addressed is how to reduce packet losses on the links. It is observed that the arrivals of the packet streams at receiving node have a bursty character. The present invention reduces the burstiness by introducing a small random delay to each packet. The delay causes the packets to jitter on the link.

Description

METHOD AND DEVICE FOR TRANSMISSION OF PACKET DATA TRAFFIC ON
A MULTIPLEXED LINK
TECHNICAL FIELD OF THE INVENTION
The present invention relates to transmission of information packets in a communication system for example a radio access network and especially to a method for transmission of packet streams and to a node handling such streams.
DESCRIPTION OF RELATED ART
UTRAN is a short name of the UMTS (Universal Mobile Telecommunication System) Radio Access Network. A mobile terminal MT, a radio base station BS and a radio network controller RNC are three basic nodes in UTRAN, and shown in figure 1. The mobile terminal MT is connected by a radio link to the radio base station BS . From the radio base station BS radio links with several mobile terminals MT are provided. The radio network controller RNC is connected to several radio base stations BS by fixed links, however in figure 1 just one exemplar of each of the nodes is disclosed. Communication in the direction to the mobile terminal MT, i.e. in the downlink direction, is set up via the radio network controller RNC and via the radio base station BS to the mobile terminal MT . The communication is also conveyed from the communication source via an exchange connected to the radio network controller RNC, however, no nodes except the radio network controller RNC, the radio base station BS and the mobile terminal MT will be described here. Communications in the opposite direction, i.e. in the uplink direction, is transmitted via the same nodes.
Figure 2 shows protocol stacks up to layer 2/3 of the radio network controller RNC, the radio base station BS and the mobile terminal MT . The protocol stack shows layers of protocols that correspond to specific functionalities for establishing communication.
Also shown in figure 1 are the fixed link interface Iub between the radio network controller RNC and the radio base station BS, and the air interface Uu between the radio base station BS and the mobile terminal MT.
The lower layers of the protocol stacks are related to physical and functional requirements of transmitting data from the sending to the receiving node. The lowest layer on both sides of the air interface Uu is thus WCDMA (Wideband Code Division Multiple Access) that corresponds to the radio multiple access technology. In the radio network controller RNC in the Iub side of the radio base station BS the transport network infrastructure LI is depicted, e.g. an ATM/AAL2 network or an IP network. Any of these is used for transferring MAC frames.
The MAC frames are produced by the above MAC (Media Access Control) layer. The MAC layer as well as the RLC (Radio Link Control) layer are located in the radio network controller RNC and in the mobile terminal MT.
The radio links shall be utilised optimally because the capacity of these radio links is limited.
Streams of information packets in the downlink direction are formatted into streams of radio frames with a certain length and format in the radio network controller RNC. The main function of the MAC layer in the downlink is to schedule the radio frames. The scheduling of the radio frames is done periodically in every 10, 20, 40 or 80 ms to adapt to the radio interface.
There are strict QoS (Quality of Service) requirements on the Iub interface. This applies also to interfaces between different radio network controllers RNC, however this interface is not shown in figure 1. The QoS requirements are strict both on delays and packet losses.
Consider a general packet communication receiving process in order to discuss the QoS requirements. Various connections 33 (shown in figure 3) are multiplexed in a queuing system, which consists of a FIFO (First In First Out) buffer 32 of size b and a server 31 with rate S. The queuing delay packets suffer when waiting in the buffer must be below a certain limit. As long as the packets of the received flows are evenly distributed over time and the total flow rate are not higher than the serving rate of the server 31, all packets will be served the moment they arrive. No queuing delays will thus arise. If, however, the traffic flow is bursty, i.e. the packets are unevenly distributed over time, while the packet flow rate is the same on a time averaged basis, packets will have to be queued in the buffer 32 before being served. The more uneven distribution of the packets, the longer the buffer 32 need to be. Packets in the buffer 32 will be delayed, the longer queue of packets the longer delay. Packets delayed longer than maximum are deleted and thereby lost.
One way of reducing the delays would of course be to increase the capacity of the server 31. Another known way to reduce the delay in the buffer, and thereby losses, is to shape the traffic according to e.g. the well known Leaky- bucket method. However, a Leaky-bucket like shaper is not applicable in UTRAN because multimedia traffic, and especially voice traffic can not be a subject of such shaping because its real-time character requires low delays. Leaky-bucket like shaping is usually applied to non-real time traffic to modify the packet arrival patterns such that it conforms to a simple, mathematically tractable traffic description. Figure 4a shows the multiplexing of packet streams on the transport network LI of fixed interface Iub. In this example there are four voice connections, each corresponding to a packet stream on one of the horizontal lines.
The multiplexing also corresponds to the order in which the packets will arrive at the fixed interface Iub. The packets are scheduled periodically with a constant interarrival time TTI (Transmission Time Interval) . In figure 4a TTIs are indicated by vertical dotted lines. The TTI are 20 ms long for voice, however, different services have different TTIs. The example with voice service on the Iub link can - be generalised to all types of services and also to other links .
When a new voice connection is established, a phase offset 42 is selected for the connection. The phase offset 42 is selected randomly, according to a uniform distribution, over the TTI. The objective of the phase offset 42 selection is to smooth the traffic and thereby decrease burstiness.
The packet transmission is started after the phase offset 42 is selected. The packet transmission can be modelled with a periodic on-off source. Packets are transmitted during the "on-periods" and not transmitted during the "off-periods" . In figures 3, 4a and 5, small boxes indicate packet transmissions in the on-periods.
Figure 5 shows four connections 53 with periodic packet arrivals multiplexed in a queuing system, which consists of a server 51 and a FIFO buffer 53. The buffer temporarily stores the arrived packets if the server 53 is busy when they arrive. The buffer depth b corresponds to the maximum delay on the fixed interface Iub as required by the QoS. If the buffer is full when a packet arrives the packet will be lost . SUMMARY OF THE INVENTION
The present invention addresses a problem that packets are lost when periodic scheduling is used on a transmission link.
The solution to this problem is based on the observation that even though a phase offset is randomly selected for each packet stream, in order to smooth the traffic, the transmissions of multiplexed packets on the link may nevertheless have a bursty character. When the packets at the end of a burst are received the receiving server is busy and the packets are delayed in the server buffer and possibly lost. Moreover, since the selected phase offset is used for the whole transmission of a stream of packets, all packets of the stream may be received in an unfavourable time with respect to the server queue.
The present invention solves the problem by a method introducing a random delay to each packet in the transmitting node. The delay is randomly selected within an interval and causes the burstiness to decrease and and the arrival order of packets of different streams to vary. Thereby the maximum server queue is reduced and, moreover, the delay will 'be more fairly distributed over the various packet streams.
The problem is also solved by a node for an electronic information system comprising an input for receiving information streams and a packet output link and that is arranged to transmit the information streams as packet streams scheduled periodically on the packet link. The transmitter is further arranged to introduce a small randomly selected delay to each packet, and thereby causing the packets to jitter on the link.
The invention is also solved by a shaper arranged to introduce a s random delay to periodically scheduled packets that are to be transmitted on a multiplexed link as streams of packet data.
The invention has the great advantage of reducing the packet losses without having to increase the receiving server capacity.
A further advantage is that losses are distributed more fairly over the streams.
A further advantage is that the invention is easy to implement with respect to the protocol stack used in communication systems such as UTRAN. The functions to introduce the random delays will then be implemented between those handled by the MAC-layer and those handled by the Transport Network Layer.
DESCRIPTION OF THE DRAWINGS
Figure 1 is a block diagram of basic nodes in a radio access system.
Figure 2 is a block diagram of UTRAN protocol layers.
Figure 3 is a block diagram showing packet streams fed to an input buffer of a receiving server.
Figure 4a is a diagram of scheduling of 4 packet streams on a link according to prior art.
Figure 4b is the same as figure 4a but with the invention applied.
Figure 5 is a diagram corresponding to that of figure 3 but with the packets scheduled periodically and time divided into transmission intervals.
Figure 6 is a flow chart showing the introduction of a random delay on the packets. Figure 7 is a time diagram showing phase relations of the packets of 4 different streams.
Figure 8 is a a block diagram of a packet data traffic shaper.
Figure 9 is a diagram showing packet losses as a function of the maximum length of the random delay introduced.
Figure 10 is a diagram showing the gain with respect to non introduced delay as a function of maximum length of the random delay introduced.
DESCRIPTION OF PREFERRED EMBODIMENTS
The present invention is based on the observation that the prior art scheduling imposes a bursty character of the packet arrivals. Figure 7 shows the prior art phase relations of 4 packets of different streams that are scheduled and multiplexed on the same link. Each packet stream corresponds to a connection, i.e. a voice connection or a multimedia connection. A packet P1-P4 from each of the stream is transmitted within the same TTI (Transmission Time Interval) . One of the packets P3 overlap in time with two other packets P3, P4. Assume that the server 51 has capacity for handling just one packet P1-P4 at the time. The third and the fourth packets P3, P4 will be buffered before being served and the fourth packet P4 will be buffered the longest time. The phase relations remain the same for subsequent TTIs, all with respect to prior art.
According to the present invention a small random delay is introduced to each packet before the packets are multiplexed on the fixed interface Iub. The phase relations between the packets will thus differ between subsequent TTIs and the burstiness will decrease. The independence of the different packet streams is maintained. Figure 4b is the same as figure 4a except for that the transmission moment for the packets P4 are slightly varying between different TTIs as a result of the random delay introduced to the packets P. The random delay interval 41 is also indicated in figure 4b.
Figure 6 is a flowchart showing the steps performed by different protocol layers of a transmitting node, in the UMTS system. In a first step 61, the packets of various streams are scheduled periodically and deterministic according to the principle disclosed with reference to figure 4a. In short this means that there is an opportunity for each stream to have a packet scheduled at a precise occasion within each TTI. The precise moment is determined by the phase offset 42 of each packet stream. The scheduling is handled by the MAC-layer of the protocol stack shown in figure 2.
In a next step 62 a small random delay is introduced to each packet P4. The delay is randomly selected from an interval 41 that is much smaller than the TTI.
In a following last step 63, the packets P4 are multiplexed on the link. This is handled by the Transport Network layer, i.e. layer 1 of the protocol stack shown in figure 2.
The small delay introduced in step 62 causes packets to jitter on the link. By jitter is meant that the arrival times differs slightly and randomly from the expected arrival times. The jitter will be in the region of [phase offset; phase offset + X] where X denotes the maximum delay.
To implement the invention just the second step 62 is added.
To perform the random-delay shaping step 62, a random-delay shaper is implemented. Figure 8 discloses the shaper 81 that comprises a random number generator 82, a table 83, a clock
85, an input from the MAC layer, an output to the transport layer and a central processor 86 that controls the functioning of the Shaper 81.
Consider downlink traffic. After MAC frames are encapsulated into Iub frames and sent towards the transport layer, they are first entering the random-delay shaper 81. The shaper 81 attaches a temporary identifier to the MAC frames and timestamps the arriving MAC frames. The random number generator 82 generates random numbers according to a uniform distribution from the interval 41 [0ms ; Xms] , X being the maximum introduced delay. A random delay is generated for each MAC frame and the random delay value is added to the timestamp of the MAC frame. The temporary frame identifiers and the related timestamps (updated with the random delay) are then inserted into the table 83, which is ordered according to the timestamp values. The MAC frames are stored in the buffer 84. If the internal clock 85 of the shaper (81) reaches the value of the first timestamp in the table (83), the MAC frame with the corresponding temporary identifier is sent to the transport network.
As a consequence of QoS requirements on maximum total delay on the fixed interface Iub, the receiving server buffer 51 depth need to be reduced correspondingly to the maximum random selected delay introduced. This could suggest that the advantages with the introduced random delay would not compensate for the disadvantage of the corresponding reduction of the receiving buffer depth 51. Consider the following example in order to evaluate the transmission performance with the invention applied:
In a voice connection, compressed voice packets are scheduled from the MAC layer every 20 ms (TTI) , they can be delayed at most 5 ms on the Iub interface, and the average packet loss should be at most around 0.1% - 1%. 387 voice connections, i.e. packet streams, are multiplexed over a 2720 kbit/sec link. The voice connections have random phase offsets 42. Voice packets are sent periodically in the 20ms TTIs according to an on-off process. The length of the on period is exponentially distributed with a mean of 10 sec. The probability of the on period is 0.5. The codec is a 12.2 kbps codec, it sends 31 bytes in the on periods. In the off periods, 8 bytes are sent in every 8th TTI.
The parameters of the voice model are not very important in our evaluation. The most important is that the voice and other services in UTRAN can be modelled as an on-off source with deterministic arrivals.
The voice packets are multiplexed in an AAL2 multiplexer and transported in ATM cells to the Base Station.
In a diagram of figure 9, the loss performance results are depicted, for various maximum lengths of the random delay applied to each packet. On the Iub interfaces there is a maximum 5ms-delay requirement. The horizontal axis of the diagram corresponds to the buffer 51 size that was decreased from 5ms to 1ms, while the uniform distribution range controlling the randomisation was increased from [0ms; 0ms] to [0ms; 4ms] . So, for example for a 3ms buffer 51, the random delay introduced in the Random Delay Shaper step 72 was chosen according to a uniform distribution over the range [0ms; 2ms] . This way, the maximum 5ms-delay requirement was met in all cases. The uniform black line shows the average loss and the dashed line shows the maximum loss for the various buffer depths.
At all 9 points in the curve, 10 simulations were run in a simulator. Each time -800000 packets were simulated.
For the maximum buffer 51 depth of 5 ms, no random delay at all was applied and the curves for 5 ms buffer depth thus represents the performance of state of the art. From figure 9 can be seen that introducing only a small delay e.g. [0ms; lms] at 4ms buffer results already in a large decrease both in the average loss, above 1% to 0.5%, and in the maximum loss, 28% to around 6%. The maximum and the average are calculated over the 10 simulation runs and over the 387 connections .
What happened is that the average delay including the randomisation delay increased, while the burstiness of the arrival's process has been reduced. This way, the losses got lower. At the same time, since the randomisation results in a constantly varying phase relation arrival order on different connections, the packet losses are distributed more fairly over the connections.
Increasing the average delay is acceptable, since the interesting performance measure is not the average but the tail probability e.g. 99% quantile of the delay.
To make it easier to quantify the advantages of the solution in the case of this example the achieved gains are showed in Figure 10. The uniform black line shows the gain related to the average packet loss (AVGGAIN) while the dashed line shows the gain regarding maximum packet loss (MAXGAIN) , compared to applying a buffer 51 depth of 5 ms and no random delay. The horizontal axis corresponds both to the buffer size and to the maximum random delay, in the respect that the maximum total delay of both shall be 5 ms . For a buffer size of 5 ms there is thus no random delay at all applied to the packets. From the curve it can be concluded that a buffer size of 3 ms and [0ms; 2ms] random delay, the maximum loss became 7 times smaller than with no random delay and 5 ms buffer 51 depth.

Claims

1. A method of transmitting (73) a number of packet data streams (S4) on a multiplexed link (53) wherein the packets (P4) of said packet data streams are scheduled periodically (71) , c h a r a c t e r i s e d in,
introducing (72) a randomly selected delay (42) to each packet (P4) and thereby causing the transmission time of the packets in said streams to jitter.
2. The method of claim 1 wherein an offset (41) is allocated to each of said packet data streams, said offset defining a moment of possible transmission of the packets (P4) of respective data stream (S4).
3. The method of claim 3 wherein said offset (42) defines the moment for transmission within transmission time intervals (TTI) .
4. The method of claim 3 wherein said randomly selected delay is smaller than the transmission time interval (TTI) .
5. The method of claim 4 wherein the said delay is selected according to a uniform distribution in a delay interval (41) [0;X] (41), X being the maximum introduced delay.
6. A method according to claim 1 wherein periodically scheduled packets (P4), and a timestamp of their respective scheduling time plus the selected random delay are buffered with reference to the corresponding packet and wherein the packets are read from the buffer in order to be multiplexed on the link when a clock time corresponds to the value of the respective timestamp plus the selected random delay.
7. The method of claim 1 or 3 wherein the scheduling is handled by a MAC protocol layer and the multiplexing by a Transport Network layer.
8. A node (RNC, BS) for an electronic information network comprising,
an input for receiving a number of information streams (S4),
an output (Iub) to a packet link (53) , and
arranged to transmit the received information streams as corresponding packet data streams (S4) scheduled periodically and multiplexed on the output link,
charcterized by,
shaping means (81) for introducing a random delay (41) to each packet (P4) and thereby cause the transmitted packets to jitter.
9. A node according to claim 8 arranged to allot a phase offset (42) to each packet data stream, said phase offset defining a possible transmission moment within transmission time intervals (TTI) .
10. A node for en electronic information system arranged to carry out the method of claim 1,2,3,4,5,6 or 7.
11. A node according to claim 8 or 10 that is a radio base station (BS) or a radio network controller (RNC) .
12. A radio access network arranged to carry out a method as claimed in claim 1, 2, 3, 4, 5, 6 or 7.
13. A radio access network comprising a node as claimed in claim 8, 9 or 10.
14. A packet data traffic shaper (81) comprising, an input for receiving streams (S4) of periodically scheduled packets (P4),
an output for transmission on a multiplexed link,
a random delay selector (82) for selecting a random delay for each packet,
a buffer (83,84) for buffering the received packets and an arrival time plus the selected delay for the respective packet,
control means (86) comprising a clock (85) and arranged to control the buffer, the delay selector and the clock and arranged to read packets from the buffer to the output when the respective arrival time plus selected delay corresponds to the clock time.
15. A node for en electronic information system including a shaper as claimed in claim 12.
PCT/SE2001/000512 2001-03-09 2001-03-09 Method and device for transmission of packet data traffic on a multiplexed link WO2002073901A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/SE2001/000512 WO2002073901A1 (en) 2001-03-09 2001-03-09 Method and device for transmission of packet data traffic on a multiplexed link

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/SE2001/000512 WO2002073901A1 (en) 2001-03-09 2001-03-09 Method and device for transmission of packet data traffic on a multiplexed link

Publications (1)

Publication Number Publication Date
WO2002073901A1 true WO2002073901A1 (en) 2002-09-19

Family

ID=20283003

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/SE2001/000512 WO2002073901A1 (en) 2001-03-09 2001-03-09 Method and device for transmission of packet data traffic on a multiplexed link

Country Status (1)

Country Link
WO (1) WO2002073901A1 (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1536598A1 (en) * 2003-11-27 2005-06-01 Siemens Aktiengesellschaft Method for packetizing time synchronous data during transmission in a packet data network
CN100438490C (en) * 2004-08-27 2008-11-26 华为技术有限公司 Average data issuing time regulating method based on statistic
US9685815B2 (en) 2014-01-16 2017-06-20 Mediatek Inc. Method for performing wireless charging control with aid of admittance detection, and associated apparatus
US9906063B2 (en) 2014-01-16 2018-02-27 Mediatek Inc. Method for performing wireless charging control of an electronic device with aid of random phase-delay packet, and associated apparatus
US10447061B2 (en) 2014-01-16 2019-10-15 Mediatek Inc. Method for performing wireless charging control of an electronic device with aid of variant slot timing and simple response indicating acknowledgement, and associated apparatus
US10554070B2 (en) 2014-01-16 2020-02-04 Mediatek Inc. Method for performing wireless charging control of an electronic device with aid of simple response indicating acknowledgement, and associated apparatus

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4412326A (en) * 1981-10-23 1983-10-25 Bell Telephone Laboratories, Inc. Collision avoiding system, apparatus and protocol for a multiple access digital communications system including variable length packets
US5850525A (en) * 1996-03-29 1998-12-15 Advanced Micro Devices, Inc. Method and apparatus for adding a randomized propagation delay interval to an interframe spacing in a station accessing an ethernet network
US5854900A (en) * 1996-05-31 1998-12-29 Advanced Micro Devices, Inc. Method and apparatus avoiding capture effect by adding a slot time to an interpacket gap interval in a station accessing an ethernet network
US6055578A (en) * 1996-08-15 2000-04-25 Advanced Micro Devices, Inc. Apparatus and method for selectively controlling transmission of consecutive packets in a network station
WO2000048367A2 (en) * 1999-02-10 2000-08-17 Nokia Wireless Routers, Inc. Adaptive communication protocol for wireless networks

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4412326A (en) * 1981-10-23 1983-10-25 Bell Telephone Laboratories, Inc. Collision avoiding system, apparatus and protocol for a multiple access digital communications system including variable length packets
US5850525A (en) * 1996-03-29 1998-12-15 Advanced Micro Devices, Inc. Method and apparatus for adding a randomized propagation delay interval to an interframe spacing in a station accessing an ethernet network
US5854900A (en) * 1996-05-31 1998-12-29 Advanced Micro Devices, Inc. Method and apparatus avoiding capture effect by adding a slot time to an interpacket gap interval in a station accessing an ethernet network
US6055578A (en) * 1996-08-15 2000-04-25 Advanced Micro Devices, Inc. Apparatus and method for selectively controlling transmission of consecutive packets in a network station
WO2000048367A2 (en) * 1999-02-10 2000-08-17 Nokia Wireless Routers, Inc. Adaptive communication protocol for wireless networks

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
SCHEDULING MONUNIFORM TRAFFIC IN A PACKET SWITCHING SYSTEM WITH LARGE PROPAGATION DELAY, IEEE TRANSACTIONS ON INFORMATION THEORY, vol. 41, no. PART 4, March 1995 (1995-03-01), HAJEK B. ET AL., pages 358365 *

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1536598A1 (en) * 2003-11-27 2005-06-01 Siemens Aktiengesellschaft Method for packetizing time synchronous data during transmission in a packet data network
WO2005062547A1 (en) * 2003-11-27 2005-07-07 Siemens Aktiengesellschaft Method for packeting time-synchronous data during transmission in a packet data network
CN100438490C (en) * 2004-08-27 2008-11-26 华为技术有限公司 Average data issuing time regulating method based on statistic
US9685815B2 (en) 2014-01-16 2017-06-20 Mediatek Inc. Method for performing wireless charging control with aid of admittance detection, and associated apparatus
US9906063B2 (en) 2014-01-16 2018-02-27 Mediatek Inc. Method for performing wireless charging control of an electronic device with aid of random phase-delay packet, and associated apparatus
US10250067B2 (en) 2014-01-16 2019-04-02 Mediatek Inc. Method for performing wireless charging control of an electronic device with aid of random phase-delay packet, and associated apparatus
US10447061B2 (en) 2014-01-16 2019-10-15 Mediatek Inc. Method for performing wireless charging control of an electronic device with aid of variant slot timing and simple response indicating acknowledgement, and associated apparatus
US10554070B2 (en) 2014-01-16 2020-02-04 Mediatek Inc. Method for performing wireless charging control of an electronic device with aid of simple response indicating acknowledgement, and associated apparatus

Similar Documents

Publication Publication Date Title
EP1374458B1 (en) Method and apparatus for implementing a mac coprocessor in a communications system
US6879561B1 (en) Method and system for wireless packet scheduling with per packet QoS support and link adaptation
EP1839412B1 (en) Interworking between cell and packet based networks
AU2003244603B2 (en) Packet Transmission Method and System, Base Station, Wireless LAN Terminal, and Wireless LAN System Using the Same
CN100407698C (en) Data transmission method for wireless link control layer
US20020093953A1 (en) System for uplink scheduling packet based data traffic in wireless system
Chebrolu et al. Communication using multiple wireless interfaces
EP0591150A4 (en) Congestion management based on multiple framing strategy.
CN102014431B (en) Anti-jittering data transmission method, equipment and system
US6963534B1 (en) Methodology for improving the performance of asynchronous data traffic over TDD/TDMA wireless networks
EP1134943B1 (en) Method of transmitting between a base station in an access network and an access network controller of a telecommunications system
JP2007506364A (en) How to empirically schedule network packets
WO2002073901A1 (en) Method and device for transmission of packet data traffic on a multiplexed link
Maqhat et al. Performance analysis of fair scheduler for A-MSDU aggregation in IEEE802. 11n wireless networks
JP2001308914A (en) Base station unit for wireless packet communication
US8355403B2 (en) Stale data removal using latency count in a WiMAX scheduler
Malkowski et al. Interaction between UMTS MAC scheduling and TCP flow control mechanisms
CN114585019B (en) ACK packet anti-congestion method, base station, terminal and system of 5G downlink TCP
CN114585018B (en) Method, base station, terminal and system for preventing ACK packet congestion of 5G downlink TCP
CN100450228C (en) Method for eliminating relativity of transmission delay between real time service and nonreal time service
Shanmuganathan et al. Performance measures of wireless protocols for ATM networks
TW421950B (en) An efficient traffic scheduling algorithm in central-control shared-media networks
Nilsson et al. A Performance Study of the XTP error control
Bada et al. Radio Resource Management in GPRS with Quality of Service
Yang et al. A signal-to-interference ratio based downlink scheduling scheme for WCDMA mobile communication system

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EE ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NO NZ PL PT RO RU SD SE SG SI SK SL TJ TM TR TT TZ UA UG US UZ VN YU ZA ZW

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): GH GM KE LS MW MZ SD SL SZ TZ UG ZW AM AZ BY KG KZ MD RU TJ TM AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE TR BF BJ CF CG CI CM GA GN GW ML MR NE SN TD TG

121 Ep: the epo has been informed by wipo that ep was designated in this application
DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
REG Reference to national code

Ref country code: DE

Ref legal event code: 8642

122 Ep: pct application non-entry in european phase
NENP Non-entry into the national phase

Ref country code: JP