JP2008061267A - Server - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To attain data transmission/reception suitable for a reception level of a portable terminal with a server even when connection is established, and further to reduce the load on a network due to useless re-transfer processing in a wireless network whose error rate is considered to be high. <P>SOLUTION: The server is connected to a wireless packet network adopting an adaptive modulation system in which the modulation system is determined based on a data transfer rate request value given from the portable terminal. The server receives from the portable terminal an error prediction value calculated based on an error rate transition at communication, and controls a re-transfer timer for re-transfer processing of packets based on the error prediction value. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a server that transmits and receives data via a best-effort wireless packet network using an adaptive modulation scheme.

In recent years, some mobile terminal devices such as mobile phones can transmit and receive data. In addition, wireless communication lines are also increasing in bit rate, and one of them is a best effort type wireless packet network.
However, in the best effort type wireless packet network, the bit error rate, the packet loss rate, the packet delay time, etc. are higher than those of the wired network due to the characteristics of wireless communication, and the cdma2000 1x-EV DO system (this is In wireless packet networks using adaptive modulation schemes such as Std.T-64 1S-2000 CS 0024 “cdma2000 High Rate Packet DataAir Interface Specification” at the Radio Industry Association ARIB) The bandwidth also has the problem of fluctuating rapidly.

  On the other hand, TCP / IP, which is an industry standard communication protocol in wired networks, has become widespread. In the case of TCP / IP, in the transport layer: TCP stack at the server on the transmission side, TCP segment transmission processing is performed based on the congestion window size value managed in the TCP stack. In a typical algorithm for determining the congestion window size value, the congestion window size starts from one segment, is doubled every time an ACK (acknowledgment response) is received from the other end on the receiving side, and within the retransmission timeout time after transmission. If no ACK is received, the congestion window size value is initialized to one segment (slow start algorithm).

In addition, the service node adopts the smaller one of the advertisement window size notified from the mobile terminal and the congestion window size according to the delivery status of the transmission data as the size of the transmission window. Is transmitted to the mobile terminal via the base station, while the mobile terminal determines the advertising window size based on the radio link radio wave intensity when receiving this TCP segment, and specifies the advertising window size in the ACK (acknowledgment) segment There are also things like sending.
JP 2002-344560 A

  However, in the TCP stack used by wire, the above slow start algorithm is effective, so the congestion window is initialized after the connection is established, and the window size is very small. Even if a high transmission rate is possible in a best-effort wireless packet network using an adaptive modulation scheme such as the cdma2000 1x-EV DO scheme, as long as there is no ACK (acknowledgment response) from another end, A data amount that does not exceed the window size is not transmitted. Therefore, even if a high transmission rate is possible, the mobile terminal requesting data transmission has to start from the lowest transmission rate, and it takes time to reach the actual high transmission rate. is there.

  In addition, even during data transmission, if re-transmission occurs due to errors that are likely to occur in the wireless network, the congestion window size is increased again from a low transmission rate to an actual transmission rate in order to reduce the congestion window size. There is a disadvantage that takes time to do. Further, in addition to the re-transmission processing control in the network layer, the transport layer: TCP corresponding to the upper layer manages a re-transmission timer based on the RTT (round trip time) value (this re-transmission timer). The value is calculated from the RTT, which is the elapsed time until receiving the ACK from the other end that is the receiving side for the segment sent from the server that is the sending side), and the retransmission algorithm in the lower layer than the transport layer Is running asynchronously. Therefore, despite the fact that the re-transmission processing is performed in the lower layer, the re-transmission timer is counted even in the upper layer TCP, and after the confirmation response is made in the re-transmission processing between the lower layer links, There is a drawback in that retransmission segments are transmitted again from the upper layer TCP, and useless packet data is transmitted to the network.

  In addition, as described in Patent Document 1, an advertisement window size is determined on the mobile terminal side based on the radio field strength of the radio link, and the transmission window size is determined based on the advertised window size and the congestion window on the service node side. In the wireless network communication where an increase in the delivery delay time can be assumed, the congestion based on the previous high communication quality even when the mobile terminal side suddenly decreases the transmission rate due to a decrease in real-time characteristics. There is a case where a window size is used, and there is a concern that an additional load is applied to the network. On the other hand, even when the transmission rate is rapidly increased, it is assumed that the congestion window size based on the previous low communication quality is used. Even in this case, there is a possibility that the network resources cannot be effectively used. In addition, consideration is not given to the above-described problem that is assumed because the retransmission process in the transport layer and the retransmission process in the lower layer are asynchronous.

  The present invention has been made in view of the above points, and enables transmission / reception of data suitable for the reception level of the mobile terminal even when a connection is established, and further has a high error rate. In such a wireless network, a server connected to a best-effort wireless packet network using an adaptive modulation method that can reduce the load on the network due to useless retransmission processing is provided.

The server according to the present invention employs the following means in order to solve the above problems.
The invention according to claim 1 is a server connected to a wireless packet network using an adaptive modulation scheme that determines a modulation scheme based on a data transfer rate request value given from a mobile terminal, and an error rate transition during communication Is received from the portable terminal, and a retransmission timer for packet retransmission processing is controlled based on the error prediction value.

  The invention according to claim 2 is the server according to claim 1, wherein when the error prediction value is high, the retransmission timer value is increased, and when the error prediction value is low, the retransmission timer is increased. It is characterized by a small value.

In order to solve the above-described problem, the present invention may be configured as a wireless packet network including a mobile terminal and a server as described below.
That is, as a configuration 1 of a wireless packet network, a best-effort wireless packet network using an adaptive modulation method that determines a connection time and a modulation method based on a data transfer rate request value from a mobile terminal connected to a wireless base station The server is connected to the Internet network and transmits a variable bit rate according to the bandwidth of the network (if the gateway located between the server and the client performs variable bit rate control) and the Internet network. For communication between mobile terminals connected to a wireless network and receiving data transferred with a variable bandwidth from the server, a TCP stack is used for the transport layer protocol in the mobile terminal and the server, Calculated based on the reception level at the mobile device Based on the transmission rate prediction value calculated based on the data transfer rate request value and the error prediction value calculated from the data reception error rate transition in the portable terminal, it is also possible to perform congestion control in the TCP stack in the server. Good.

  Further, in the wireless packet network of Configuration 1, a predicted transmission rate at which the server can transmit data, calculated from a reception level transition at the mobile terminal that is receiving data, and a data reception error rate at the mobile terminal A configuration may be adopted in which the congestion window size value managed in the TCP stack of the server that performs transmission is determined by notifying the server of the error prediction value calculated from the transition.

  Further, by notifying the server of the configuration 1 of the error prediction value calculated from the data reception error rate transition in the mobile terminal of the configuration 1, the retransmission timer value managed in the TCP stack in the server It is good also as a structure which controls.

  In addition, regarding the notification of the transmission rate prediction value and the error prediction value transmitted by the mobile terminal having the configuration 1, the server notifies the server only when there is a change in the prediction result. It is good also as a structure which uses.

  Further, the mobile terminal that is the data receiving side from the server is adapted according to the result of the reception level, the wireless network interface unit that controls data communication, the reception level quality monitoring unit that monitors the reception level with the wireless network, and A data transfer rate request calculation unit for calculating a data transfer rate, a transmission rate prediction unit for predicting a transmission rate suitable for the reception level from the result of the data transfer rate request calculation unit, and an error for calculating an error rate for received data A rate calculation unit, an error rate prediction unit that predicts an error rate based on a result of the error rate calculation unit, a transmission rate value predicted by the transmission rate prediction unit, and an error prediction value predicted by the error rate prediction unit It is preferable that it is comprised with the notification means for notifying to a server.

As described above, in the wireless packet network, the congestion control in the TCP stack of the server is performed based on the transmission rate prediction value based on the data transfer rate request value calculated based on the reception level at the mobile terminal and the error prediction value. Do. In this case, the congestion window size value is calculated from the transmission rate prediction value that can be transmitted by the server and the data reception error rate transition in the mobile terminal, calculated from the reception level transition in the mobile terminal that is receiving data. This is based on the error prediction value.
In addition, the error prediction value calculated from the data reception error rate transition in the mobile terminal is notified to the server, thereby changing the retransmission timer value managed in the TCP stack of the server.

  According to the present invention, in TCP / IP communication in which a connection is established via a wireless network, the congestion window size value suitable for the reception level at the mobile terminal on the receiving side is obtained at the time of connection establishment and before and after the occurrence of an error. Based on this, since the server on the transmission side can send data, it becomes possible to effectively utilize the network resources in the best-effort wireless packet network using the adaptive modulation method with large bandwidth fluctuation.

According to the present invention, the following effects can be obtained.
According to the present invention, in a best-effort wireless packet network using an adaptive modulation method with a large bandwidth variation, even when a connection is established, data transmission / reception suitable for the reception level of the mobile terminal connected to the wireless packet network is achieved. Can be performed with the server. Further, even in a wireless network that is said to have a high error rate, it is possible to reduce the network load by useless retransmission processing.

Hereinafter, the present invention will be described based on one embodiment.
FIG. 1 shows a connection time between a mobile terminal (100) and a radio base station based on a data transfer rate request value notified by the mobile terminal (100) using the server (102) according to an embodiment of the present invention. 1 is a conceptual diagram illustrating a configuration example of a best-effort wireless packet network using an adaptive modulation scheme that determines a modulation scheme.
The mobile terminal (100) in FIG. 1 is equipped with a communication function for performing communication according to TCP / IP on a wireless network using an adaptive modulation method, and wireless data with the wireless base station (101). While performing link layer communication, the TCP segment is transmitted from the server (102) via the IP network (103) and the wireless base station (101).

FIG. 2 is a block diagram showing a hardware configuration of the mobile terminal (100).
In FIG. 2, 200 is a wireless network interface unit for controlling data communication, 201 is a reception level quality monitoring unit for monitoring the reception level with the wireless network, and 202 is a required value of the optimum data transfer rate according to the result of this reception level. Is a data transfer rate request calculation unit that calculates the transmission rate, 203 is a transmission rate prediction unit that predicts a transmission rate suitable for the reception level from the result of the data transfer rate request calculation unit (202), and 204 is an error rate for the received data. An error rate calculation unit 205 is an error rate prediction unit that predicts and calculates an error prediction value based on the result of the error rate calculation unit (204), and 206 is a transmission rate value predicted by the transmission rate prediction unit (203). And a TCP segment for generating a TCP segment including the error prediction value predicted by the error rate prediction unit (205) A tacking unit, 207 is a timer unit used to notify a time interval for calculating the average value from the required data transfer rate value and a time interval for calculating an error prediction value from the error rate, and 208 is for each packet slot. Requested data transfer rate, average value calculated from this data transfer rate required value, predicted transmission rate calculated from this average value, error rate calculated for each packet slot, this error rate It is a memory part for storing data, such as an error prediction value calculated from.

  When the cdma2000 1x-EV DO method is used as the adaptive modulation method, this method uses extremely accurate data communication obtained by taking into account predictions and corrections based on statistical data such as error rate of past downlink data transmission. The terminal side is provided with a predicted downlink data communication rate (DRC; Data Rate Control Bit) indicating the rate directly as a table, and based on this table, the predicted downlink data communication rate is notified from the terminal to the base station. However, the predicted downlink data communication speed is used as the data transfer rate request value.

Below, operation | movement by the said structure is demonstrated using the flowchart shown to FIGS.
The reception level quality monitoring unit (201) in the portable terminal (100) measures the reception level in the wireless network with the wireless base station (101) (S300).
The data transfer rate request calculation unit (202) receiving the reception level measured from the reception level quality monitoring unit (201) calculates a data transfer rate request value (S301). The measurement and calculation in S300 and S301 are performed for each packet slot in the network layer. Further, the calculated data transfer rate request value is stored in time series in the memory unit (208) in the portable terminal (100) (S302).

  Next, a time series stored in the memory unit (208) in response to an interrupt request from the timer unit (207) in the portable terminal (100) that generates an interrupt at predetermined fixed time intervals (synchronized with each packet slot) A required data transfer rate request value is read (S303), and an average value of the requested data transfer rate values for a predetermined time is calculated (S304). In general, it is known that the congestion window size varies for each RTT value in the TCP used for normal wired communication. Therefore, it is desirable to use the RTT value in the same manner as a fixed time. Further, the calculated average value is stored in time series in the memory unit (208) in the portable terminal (100) (S305). Here, with the passage of time, the average value is stored in the memory unit (208).

Next, using the average value stored in the memory unit (208), the transmission rate prediction unit (203) predicts the transmission rate (S306). Although the outline of the prediction method is demonstrated below, the prediction method used by the following description is an example, It does not limit to this.
The average values calculated from the required data transfer rate values stored in the memory unit (208) are a and b in the order of time, and the value predicted from the average value is c.
When the values of a and b are compared, it is assumed that “a <b”. In this case, unless “b” is the upper limit value of the fluctuating bandwidth, the transmission rate value “c” is expected to be “b <c”. Therefore, the predicted transmission rate value “c” is “b− It can be predicted based on the transition at “a” (bandwidth expansion) (for example, c = b + (b−a), but can be obtained as the upper limit or less).

Now assume the case of “a> b”. In this case, unless “b” is the lower limit value of the network, the predicted transmission rate value “c” is expected to be “b> c”, so that this transmission rate is “b−a” (at this time) It can be predicted based on a transition (bandwidth reduction) at a negative value (for example, c = b + (ba), but can be obtained as a lower limit value or more). The predicted value calculation from these data transfer rate request values is performed regardless of the presence / absence of data actually transmitted / received between the portable terminal (100) and the server (102), and a new transmission is made to the memory unit (208). It updates as a rate prediction value (S307).
The transmission rate is predicted as described above.

  Next, the case where the portable terminal (100) establishes a TCP connection with the server (102) and transmits / receives data will be described.

  First, negotiation with the server (102) is started by a call from the mobile terminal (100). In this case, the mobile terminal (100) transmits a TCP ACK segment to be transmitted at the time of negotiation in a segment including the latest transmission rate prediction value stored in the memory unit (208) in the header or payload field of the TCP segment. Send. The server (102) that has received the predicted transmission rate uses the predicted transmission rate received from the portable terminal (100) as the congestion window size value managed by the internal TCP stack. Therefore, in the server (102), even when there is no ACK reception from the mobile terminal (100), it is possible to transmit the data amount up to the predicted transmission rate for the TCP segment transmitted to the mobile terminal (100).

  Next, the mobile terminal (100) that has started to receive the TCP segment from the server (102) performs the above-described transmission rate prediction calculation and obtains error packet information from the lower layer to obtain the network that is the lower layer. The error rate calculation in the layer is performed by the error rate calculation unit (204) (S308). The calculation is performed for each packet slot in the network layer as in the data transfer rate request calculation unit (202).

  Next, the value of the error rate obtained here is stored in the memory unit (208) in time series (S309). Further, from the error rates stored in time series, the error prediction value is calculated at every certain time interval due to the interruption from the timer unit, similarly to the transmission rate prediction performed by the transmission rate prediction unit (203) (S310). Here, the error rate is read from the memory unit (208), and the average value is calculated as an error prediction value and stored in the memory unit (208).

  On the other hand, the portable terminal (100) is based on the average value obtained from the data transfer rate request value when performing a transmission rate prediction calculation at the time of establishing a connection for performing data transmission / reception with the server (102). The transmission rate prediction value used for the congestion window size at the server is the calculated transmission rate prediction value minus the data rate corresponding to the error prediction value and the header information in the lower layers including the TCP layer. Then (S311), it is transmitted and notified to the server in the form included in the header or payload field in the TCPACK segment from the portable terminal (100) (S312). As described above, the transmission rate prediction value used for the congestion window size in the server is corrected based on the error prediction value so as to reduce the value, thereby suppressing the occurrence of errors.

Further, the error prediction value calculated by the portable terminal (100) is notified to the server (102), thereby changing the retransmission timer in the TCP stack of the server (102).
For example, when sending a TCP segment, the server (102) that has received a notification with a high error prediction value sets the retransmission timer in this TCP segment to the normal “N times (N is a number of 1 or more)”, In the server (102) that has received the notification with a low error prediction value, the retransmission timer in this TCP segment is set to “M times the normal number (M is a number of 1 or less)”, so that unnecessary retransmission packets are transmitted on the network. Reduces sending to

  As described above, according to the present invention, in TCP / IP communication in which a connection is established through a wireless network using an adaptive modulation method, the mobile terminal on the receiving side is able to establish a connection and before and after an error occurs. Effective use of network resources in a best-effort wireless packet network that uses an adaptive modulation method with large bandwidth fluctuations because the sending server can send data based on the congestion window size value suitable for the reception level It becomes possible to do.

It is a structural example of the radio | wireless packet network in one embodiment of this invention. It is a block diagram of the portable terminal used with the radio | wireless packet network in one embodiment of this invention. It is a process flowchart in the portable terminal in one embodiment of this invention. It is a process flowchart (continuation of FIG. 3) in the portable terminal in one embodiment of this invention. It is a process flowchart (calculation of an error prediction value) in the portable terminal in one embodiment of this invention.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 100 ... Portable terminal 101 ... Wireless base station 102 ... Server 103 ... IP network 200 ... Wireless network interface part 201 ... Reception level quality monitoring part 202 ... Data transfer rate request | requirement calculating part 203 ... Transmission rate prediction part 204: Error rate calculation unit, 205 ... Error rate prediction unit, 206 ... TCP stack unit, 207 ... Timer unit, 208 ... Memory unit

Claims (2)

A server connected to a wireless packet network using an adaptive modulation scheme that determines a modulation scheme based on a data transfer rate request value given from a mobile terminal,
A server that receives an error prediction value calculated from an error rate transition during communication from the mobile terminal and controls a retransmission timer for packet retransmission processing based on the error prediction value.   2. The server according to claim 1, wherein when the error prediction value is high, the retransmission timer value is increased, and when the error prediction value is low, the retransmission timer value is decreased.

Images