CN1215679C - Method for controlling flow of data transmisison - Google Patents

Abstract

The present invention relates to a method for controlling the flow of data transmission, wherein data packets are sent to FIFO or RAM processed by downstream receiving equipment to be slowly stored by upstream transmission equipment; the input transmission rate is controlled on an interface of FIFO/RAM; the input end is provided with a time scale; the flow of statistic data is calculated according to the time scale. When the flow of the data reaches or exceeds the preset flow limit rate, the redundant data packets are discarded. Meanwhile, the output transmission rate is controlled by FIFO/RAM so as to conform to the requirements of the downstream receiving equipment. The flow control is realized, the resources of slow storage are simultaneously saved, and thereby, the design cost is reduced. The present invention can be widely applied to the fields of data transmission and communication.

Description

Flow Control Method for Data Transmission

Technical field:

The present invention relates to a flow control method, more specifically, to a flow control method in programmable logic design.

Background technique:

At present, a commonly used flow control technology is a flow control technology similar to the store-and-forward form shown in Figure 1, that is to say, the data packets with an uncertain rate are input from the ingress to the FIFO with a predetermined size. In the memory (FIFO), according to the processing capability of the user, the output of the FIFO is transmitted to the downstream device at a set fixed rate. In this design, FIFO should be made as large as possible to reduce the loss of data packets during bursts. When the number of incoming data packets exceeds the capacity of the FIFO due to a burst, the excess data packets will be discarded. Therefore, the capacity of the FIFO directly affects the reliability of data packet transmission.

Those of ordinary skill in the art know that the size of the FIFO directly affects resource utilization from another aspect. The larger the FIFO, the greater its resource consumption. From the perspective of saving resources, the smaller the FIFO, the better; and from the perspective of reliability of transmitting data packets, the larger the FIFO, the better. The contradiction between these two aspects makes the design and use of FIFO in flow control difficult to grasp and control. In fact, the downlink device in the process of data transmission often has a cache function, which can be directly used. Based on this current situation, the present invention considers canceling the FIFO in the process of data transmission, and utilizing the storage function of the downlink device itself. Such an approach can achieve the purpose of saving resources to the greatest extent on the premise of satisfying the realization of flow control.

Invention content:

In order to overcome the deficiency that the commonly used flow control method consumes more resources, the present invention provides a flow control method. In most cases (preferably, the downlink processing device itself has a cache performance of a certain scale), A flow control method using time-scale statistics. Implementing this method can not only better achieve the purpose of flow control, but also save a large amount of storage resources used for caching, thereby achieving the effect of reducing design costs.

The flow control method is easy to implement and operate, and only needs to control the input or output transmission rate on one interface. When the data flow reaches or exceeds a certain preset flow rate, redundant data packets will be discarded, wherein the flow limit rate can be set according to the total data packet traffic in transmission. The current limiting mechanism can be controlled serially, data packets can be continuously executed with different current limiting rates, and more than 20 different current limiting rates can be constructed on one interface. The method has a wide range of applications, and can be used on a network switch, can also be used on a network server for flow control, and can also be used on input and output interfaces and router interfaces.

In order to achieve the above object, a flow control method for data transmission is designed, the method includes the following steps:

The uplink data transmission device sends the data packet to a FIFO or random access memory (RAM) for buffering;

The FIFO/RAM controls the input transmission rate on an interface, and when the data flow reaches or exceeds a certain preset current limiting rate, redundant data packets will be discarded;

At the same time, the FIFO/RAM controls its output transmission rate to meet the needs of the downlink receiving device;

Wherein, a time scale is set in the FIFO/RAM, and the data flow is counted with reference to the scale. When the data flow is greater than a certain preset value, the data packets entering after this time period will be discarded. Through the Statistical and discarding methods to achieve flow control.

The FIFO/RAM is the FIFO/RAM of the downlink receiving device itself, rather than additional hardware resources.

Considering that most of the devices in the system have their own buffer memories, the present invention first designs a time scale when implementing flow control. The length of the scale can be determined according to the buffer capacity of the downlink device, and the data flow is counted with reference to the scale. Of course, this kind of traffic statistics can be classified and counted according to the data type, so that classified traffic control can be realized. The fact that the data flow rate is greater than the preset value also includes many situations. One is that the number of data packets is greater than the preset value, and the other is that the effective number of bytes of the data packet is greater than the preset number of bytes. Incoming packets after this time period will be dropped.

The implementation of the present invention realizes flow control through the method of time scale statistics and saves resources for buffering, thereby reducing design cost.

Description of drawings:

FIG. 1 is a schematic diagram of a flow control method using FIFO in the prior art.

Fig. 2 is an illustration of time-stamped data packet current limiting using the method of the present invention.

Fig. 3 is an illustration of time-stamped data byte count current limiting using the method of the present invention.

Detailed ways:

A more complete understanding of the present invention will be obtained from the following detailed description and accompanying drawings of embodiments of the present invention. Of course, the accompanying drawings should not be used to limit the present invention to specific examples, but are only used to explain and help understanding.

In Figures 2 and 3, T is the time scale length, and this parameter can be set arbitrarily according to the cache and processing capabilities of the downlink device. If the throughput capability of the downlink device is relatively strong, you can set the value of T slightly larger, otherwise it needs to be set slightly smaller. 1, 2, 3, 4, 5 are the data packet signs input by the device, 1 represents the first data packet, 2 represents the second data packet, 3 represents the third data packet, and so on, n1, n2, n3, n4, n5... are the effective byte length of each data packet.

The present invention is realized by FPGA (Field Programable Gate Array Field Programmable Gate Array) logic. Data packets can be classified and processed according to needs, that is, different types of data packets are assigned different priorities, and statistics, analysis, and processing are performed separately. For each type of data message, one of the following methods can be used to process it.

In Fig. 2, when the time scale counter is zero, the input data and output data statistics counters are cleared, and new statistics are started at the same time. Because it is set to output 3 data packets within the time T-this value is preset, and different values can be set according to the processing capacity of the downlink device-so, after the fourth input in this time scale The data packets will be discarded, so as to achieve the purpose of current limiting and flow control.

In another embodiment shown in FIG. 3 , the counting of data does not take the number of data packets as a reference, but counts the number of bytes of data. Because it is set to output n bytes of data within the time T—this value is preset, and different values can be set according to the processing capacity of the downlink device. Here, it is assumed that n1+n2<n<n1+ n2+n3 - Therefore, when the output of the third data packet ends, the output byte has exceeded the set value n, and the data packets after the fourth input within this time scale will be discarded. At the same time, the maximum value of the last data packet passed is also set in the logic. When the length of the last passed data packet is greater than the set value, the output of this data packet will be suspended and the error flag will be set at the same time, thus preventing Because of the impact of the last data packet's overlength on the downlink device.

Both of the above two current limiting methods avoid the mechanism of using a large-capacity FIFO, thereby saving resources and reducing costs.

According to the above description, those skilled in the art can easily implement the method of the present invention. The above-described invention may, of course, be embodied in other specific forms without departing from the spirit or essential characteristics as disclosed. It is therefore to be understood that the invention is not to be limited by the foregoing illustrative details.

Any reference signs in the description do not limit the protection scope of the present invention. The word "comprising" does not exclude the presence of other components than those listed in the present invention. "A" preceding a component or unit does not exclude the presence of a plurality of such components or units.

Claims (5)

1. A flow control method for data transmission, said method comprising the steps of: The uplink data transmission device sends the data packet to a first-in-first-out memory or random access memory for buffering; The first-in-first-out memory or random access memory controls the input transmission rate on an interface, and when the data flow reaches or exceeds a preset current limiting rate, redundant data packets will be discarded; At the same time, the first-in-first-out memory or random access memory controls its output transmission rate to meet the needs of the downlink receiving device; Wherein, the input end of the first-in-first-out memory or random access memory is provided with a time scale, the length of the time scale can be determined according to the caching capacity of the downlink device, and the data flow is counted with reference to the scale, when the data flow is greater than the preset When the value is set, the data packets entering after this time period will be discarded, and the flow control will be realized through the methods of counting and discarding; The first-in-first-out memory or random access memory is the first-in-first-out memory or random access memory of the downlink receiving device itself, rather than additional hardware resources. 2. The method according to claim 1, wherein when counting data traffic, classification and statistics can be performed according to data types, so as to realize classification traffic control. 3. The method according to any one of claims 1-2, wherein said when the data flow rate is greater than a preset value means that the number of data packets is greater than a preset value. 4. The method according to any one of claims 1-2, wherein when the data flow is greater than a preset value, it means that the number of effective bytes of the data packet is greater than the preset number of bytes. 5. The method according to claim 1, characterized in that the current limiting method is serially controlled, data packets can be continuously executed with different current limiting rates, and several different current limiting rates can be constructed on one interface Current limiting rate.

Images