JPS63151125A - Packet multiplexing system - Google Patents

Abstract

PURPOSE:To increase the number of packets which can be transferred per unit time, by modulating respective parallel signals consisting of packets read out from storage circuits by light beams different in wavelength to multiplex signals with respect to wavelength and transmitting them to an optical communication line. CONSTITUTION:Packets coming from input communication lines IN1-INs are converted to an optional number (n) of bits by series-parallel converting circuits 101-10s and these bits are inputted to and stored in storage circuits 111-11n. Packets stored in storage circuits 111-11n are read out when no packets are transmitted to optical communication lines OUT1-OUTt. A direction to an objective optical communication line is selected by selecting circuits 161-16n, and packets are converted to optical signals by electrooptic converting circuits 121-12n which emit light with wavelengths different from each other, and they are multiplexed by multiplexers 131-13t with respect to wavelength and are transmitted to optical communication lines OUT1-OUTt. Thus, the packet transfer time to optical communication lines is shortened without increasing the operation speed of circuits, and the number of packets which can be transferred per unit time is increased.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a so-called store-and-forward packet switching device that temporarily stores packets arriving from a plurality of input communication paths and transfers them to a single optical communication path. A local area network (LA) that uses a multipoint access method in which a single optical communication channel is time-division multiplexed with multiple packet transmission circuits.
The present invention relates to a packet multiplexing method in N).

[Conventional technology]

FIG. 4 shows an example in which the conventional packet multiplexing method is applied to a packet switch. In FIG. 4, 101 to 10s are serial/parallel conversion circuits, 1) 1 to lln are storage circuits, and 141 to
14t is a parallel-to-serial conversion circuit, 151 to 15t are electrical/optical conversion circuits, and 161 to 16n are selection circuits. Also, IN, ~IN, are human communication channels, and 0IJTI~0U
Tt is an optical communication channel.

Packets arriving at each of the plurality of input communication paths IN, ~IN, are transferred to arbitrary optical communication paths OUT, ~0UTt, by selection circuits 161 to 16n. here,
Storage circuits 1) 1 to lln form a queue to prevent two or more packets from colliding with each other when output to the same optical communication channel at the same time.

Next, the operation of this conventional example will be explained. Input communication path I
The packets input from N, ~IN, are converted into arbitrary bits (n, ) by the serial/parallel conversion circuits 101 to Iln, respectively.
The write/read operation speed conditions of the storage circuits 1) 1 to lln are relaxed.

Storage circuit 1) 1 to lln are queues (QUEtlE)
is configured, and when a packet exceeding the amount of data that can be transferred on the optical communication path 0tlT, ~OUT, arrives,
It is installed to temporarily store packets and send them out at different times, as well as convert input and output speeds and output packets at any speed. Optical communication path 01JTI~0
During the time when packets are not being sent to each of the UTts (7Q), the packets accumulated in the storage circuits 1 to 11n are read out, and the selection circuits 161 to 16n select the route toward the target optical communication path. Parallel-serial conversion circuit 141
After the packets stored in parallel are converted into serial by ~14t, they are converted into optical signals by electrical/optical conversion circuits 151-1st and sent to optical communication paths OUT and ~0UTL, respectively.

[Problem that the invention seeks to solve]

In such a conventional multiplexing system, the transmission speed of the optical communication path is mainly limited by the operating speed of the parallel/serial conversion circuits 141 to 14t. Therefore, in order to shorten the packet transfer time to the optical communication path and increase the number of packets that can be transferred per unit time, it is necessary to increase the speed of the parallel-to-serial conversion circuit, but the parallel-to-serial conversion circuit is complicated. However, there was a drawback that it was difficult to increase the speed and became a bottleneck.

An object of the present invention is to provide a packet multiplexing method that improves the number of packets that can be transferred per unit time by eliminating the above-mentioned drawbacks.

[Means for solving problems]

The present invention converts packets arriving from a plurality of input communication paths into arbitrary bits into serial/parallel data, stores them sequentially in a parallel state in a plurality of storage circuits, and converts the packets arriving from a plurality of input communication paths into arbitrary bits at a time when no packets are being transmitted on the optical communication path. In a packet multiplexing method in which packets are read out from the storage circuit, multiplexed and transmitted over the optical communication path, each of the parallel signals consisting of the packets read out from the storage circuit is modulated with light of a different wavelength. It is characterized by wavelength multiplexing and sending out to the optical communication path.

Further, in the present invention, it is preferable that the different wavelengths are set near the zero dispersion wavelength of the optical communication path.

[For production]

The operation of the present invention will be explained with reference to explanatory diagrams shown in FIGS. 3(a) and (hl). Data is a bucket of optical signals with the same wavelength λ) A, B, C
are transferred serially through the same optical communication channel in this order. In contrast, in the present invention, optical signals modulated with light of different wavelengths for the parallel signals selectively read out from the storage circuit are wavelength-multiplexed by a multiplexer and transferred over an optical communication path.

FIG. 3(b) shows this transfer state, in which packets A, B, and C are each divided into packets AI with wavelength λ1.
.

B, ,C, and packets A2, B2, C2- with wavelength λ2
..., A, , B, , C, of wavelength λ7 are multiplexed and transferred in parallel on the same optical communication path.

Now, using queuing theory and assuming that the packet length follows an exponential distribution, the average delay time from when a packet arrives at the multiplexer to when it completes transfer to the optical communication channel is T I in the conventional example. In the case of the present invention, Tt is given by the following equations.

Here, P is the average packet length, V is the transmission speed of the conventional optical communication path, v2 is the transmission speed of the optical communication path of the present invention, ρ is the usage rate of the optical communication path, n is the number of wavelength multiplexing, and the above ( As is clear from Equations 1) and (2), if the transmission speeds of the optical communication channels are equal (vI=v2), the average delay time of the present invention is 1/1/2, where the number of wavelengths multiplexed is n, compared to the conventional example.
If n is shortened or the average delay time can be made equal ("r+ = T!), the transmission speed of the optical communication channel can be reduced to 1/n in the present invention.

Therefore, it is possible to increase the number of packets that can be transferred per unit time.

In addition, by setting the wavelength of light near the zero-dispersion wavelength of the optical communication path, the phase difference between multiple optical signals transmitted in parallel can be reduced, allowing long-distance transmission (10 km or more) or high-speed transmission (IGbit/ s or more) is possible.

〔Example〕

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram of a first embodiment of the present invention in which the present invention is applied to a packet switch.

The present embodiment includes eight serial-to-parallel converter circuits 101 to 10s to which input communication paths IN and ~IN are respectively connected and which convert human-powered packets into serial-to-parallel signals into arbitrary bits (n) and output the data.
Four of them are connected to each of the serial/parallel conversion circuits 101 to 10s, forming a queue, and when more packets than the amount of data that can be transferred via the optical communication path arrive, the packets are temporarily stored and sent out at a different time. , configured to perform speed conversion of human output and send packets at an arbitrary speed,
n storage circuits 1) 1 to lln that read out packets stored in storage circuits 1) 1 to lln at times when no packets are being sent to each optical communication path 01JT+ to 0IJT, and this storage circuit 1) 1 n selection circuits 161 to 16n that select and output a route toward the target optical communication path for the packet read from ~lln, and each selection circuit 1
n electrical/optical conversion circuits 121 to 12n connected to the respective terminals 61 to 16n, each of which modulates a packet with a different wavelength of light, converts it into an optical signal, and outputs an optical signal; The wavelength-multiplexed optical signals are wavelength-multiplexed and each wavelength-multiplexed signal is connected to one optical communication path OUT l~
It includes t multiplexers 131 to 13t that output at 0tlTt. In addition, here, the number of multiplexers and optical communication channels t
The reason why S is shown differently from the number S of input communication paths is that the number of outputs does not necessarily match the number of inputs due to speed conversion in the storage circuit.

Comparing this embodiment with the conventional example shown in FIG. Electrical/optical conversion circuits 121 to 120 + multiplexer 13
It can be seen that this is due to the configuration of 1 to 13t.

Next, the operation of this embodiment will be explained. Input communication path I
The packets arriving from N and -IN are serial-parallel converted into arbitrary bits (n) by serial-parallel conversion circuits 101 to 10s, respectively, and input and stored in storage circuits 1)1 to lln, respectively. The packets stored in the storage circuits 1) 1 to 1) n are read out at a time when no packets are being sent to each of the optical communication paths OUT+ to 0UTt, and sent to the target optical communication path by the selection circuits 161 to 16n. A direction is selected, which is converted into an optical signal by electric/optical conversion circuits 121 to 12n that emit light at different wavelengths, and wavelength multiplexed by multiplexers 131 to 13t to form an optical communication path 0.
1JT+ to 0tlT.

FIG. 2 is a block diagram of a second embodiment of the present invention in which the present invention is applied to a LAN. The second embodiment replaces the selection circuits 161 to 16n in the first embodiment shown in FIG. circuit 31~3s
The other parts are the same as in FIG. However, the number of input communication channels and the number of outputs are the same. In LSN, one optical communication path 01JT is time-divisionally used by a plurality of distributed packet multiplexers 1) to ls, so the usage status of each packet multiplexer 1) to 1s and the optical communication path OUT is is monitored by the monitoring circuits 31 to 3s, and at a time when no packets are being sent out on the optical communication path OUT, the packet multiplexing device in which the packets are stored in the storage circuits 1) 1 to lln sends the packets to the optical communication path. do.

Storage circuit 1) Successive output control of 1 to lln is performed by monitoring circuit 31 to
This embodiment differs from the first embodiment shown in FIG. 1 only in that it is performed by step 3s, and the other operations are the same.

In the above embodiment, in order to reduce the phase difference between multiple optical signals transmitted in parallel and enable long-distance transmission (1 km or more) or high-speed transmission (I Cbit/s or more), , the wavelength is set near the zero dispersion wavelength of the optical communication channel.

〔Effect of the invention〕

As explained above, the present invention can reduce the time required to transfer packets to an optical communication path without increasing the operating speed of the circuit, and has the effect of increasing the number of packets that can be transferred per unit time. . Therefore, according to the present invention, high multiplex transmission of packets can be achieved using circuits with relatively slow operating speeds.
It is possible to increase the capacity of the optical communication path, and it also reduces the delay time from the time a packet arrives at the multiplexer until the transfer to the optical communication path is completed. High-speed data transmission becomes possible, and the effect is significant.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of a first embodiment of the present invention in which the present invention is applied to a packet switch. FIG. 2 is a block diagram of a second embodiment of the present invention in which the present invention is applied to a LAN. FIGS. 3(a) and (bl) are detailed explanatory diagrams of the present invention. FIG. 4 is a block configuration diagram when the conventional method is applied to a packet switch. 1) to 1s...packet multiplexing device, 31~3S...
・Monitoring circuit, 101~10s...Serial-to-parallel conversion circuit, 1
)1-1)0...Storage circuit, 121-12n, 1
51~15t...Electrical/optical conversion circuit, 131~13t
...Multiplexer, 141-14t...Parallel-serial conversion circuit,
161-16n...Selection circuit, IN, ~IN,...
Input communication path, OUT, OUT, ~0UTt... optical communication path, λ. ~λ7...Wavelength. Patent applicant: Nippon Telegraph and Telephone Corporation, -1 agent
Patent Attorney Nao Ide Takashi Procedural Amendment April 8, 1988 1, Case indication 1988 Patent Application No. 298349 2, Title of invention Packet multiplexing method 3, Relationship with the person making the amendment Patent application Address 1-1-6 Uchisaiwai-cho, Chiyoda-ku, Tokyo Name (422) Nippon Telegraph and Telephone Corporation Representative Makoto Fuji
Kou 4, Agent 5, i: Date of Ii construction order (voluntary amendment), 9-4-8, Contents of amendment fil Formula (2) % formula % ( on page 6, line 10 of the specification)
2) Correct as follows. (2) Page 8 of the specification, lines 8 to 9, "n storage circuits 1) 1 to lln that read out the packets stored in storage circuits 1) 1 to lln," are replaced with "accumulated packets. n storage circuits from which 1) 1 to ll are read out
Correct it with "n". (3) "Readout control" on page 10, line 17 of the specification is corrected to "readout control".

Claims (2)

[Claims] (1) Packets arriving from multiple input communication paths are serial-parallel converted into arbitrary bits and stored in parallel in multiple storage circuits, and the above-mentioned storage is performed during a time when no packets are being transmitted on the optical communication path to be sent out. In a packet multiplexing method in which packets are read out from a circuit, multiplexed onto the optical communication path, and transmitted, each of the parallel signals made up of packets read out from the storage circuit is modulated with light of a different wavelength. A packet multiplexing method characterized in that the packets are multiplexed and sent to the optical communication path. (2) The packet multiplexing method according to claim (1), wherein the different wavelengths are respectively set near the zero-dispersion wavelength of the optical communication path.