Network latency recovery for internet gateways

1. Field of the Invention

The present invention generally relates to network latency, and more particularly to network latency recovery for Internet gateways.

2. Description of the Related Art

T.38 is an emerging International Telecommunications Union (ITU) standard for fax transmission over Internal Protocol (IP) networks. The standard defines how Group 3 end-to-end facsimile terminals can connect to each other through the Internet. Fax has traditionally been transmitted over standard telephone lines. Transmitting fax over a data network presents new opportunities for enhanced fax delivery, such as real time fax.

The ITU-T T.38 recommendation, which is incorporated herein by reference, indicates that support for V.34 is for further study. V.34 is a standard for modems operating at data signalling rates of up to 33,600 bit/s.

Briefly, a communications system includes an internet protocol network, a calling modem and an answering modem configured to effectively handle network latency. The calling modem includes a timer storing a network latency value. The network latency value is compared to a network latency threshold. If the network latency value is less than the network latency threshold, then a high speed modem connection is continued. If the network latency value is greater than the network latency thershold, then the high speed modem connection is terminated. In place of the high speed modem connection, a low speed modem connection is established. The calling modem is operable to transmit a modem connection selection signal for selecting a high speed modem connection or a low speed modem connection. The answering modem is operable to receive the modem connection selection signal.

A better understanding of the present invention can be obtained when the following detailed description of the preferred embodiment is considered in conjunction with the following drawings, in which:

FIG. 1 is a diagram showing exemplary signalling between a calling side and an answering side of a communications system with an acceptable network latency;

FIGS. 2A–2B provide a flow chart of an exemplary network latency recovery process in accordance with the signalling and communications system of FIG. 1 ;

FIG. 3 is a diagram showing exemplary signalling between a calling side and an answering side of the communications system of FIG. 1 with an unacceptable network latency; and

FIG. 4 is a block diagram of a calling modem in a calling gateway of FIGS. 1 and 3 .

Turning now to the drawings, FIG. 1 shows a communications system C including end-point facsimile terminals F 1 and F 2 and Internet T.38 gateways G 1 and G 2 . F 1 and G 1 constitute the calling side of the system C, and F 2 and G 2 constitute the answering side of the system C. The gateway GI includes a calling modem, and the gateway G 2 includes an answering modem. The facsimile gateways G 1 and G 2 allow for the facsimile terminals F 1 and F 2 to communicate transparently through an internet protocol (IP) network such as the Internet. Communication between the gateways G 1 and G 2 , for example, can include transmitting a facsimile. With respect to facsimile communications, the calling gateway G 1 encodes the fax signals into IP fax packets for transmission, and the answering gateway G 2 decodes the transmitted IP fax packets back into fax signals. FI and GI can be embodied in an Internet-aware facsimile device, and F 2 and G 2 can be embodied in another Internet-aware facsimile device. F 1 is shown sequentially providing a CNG signal sequence, a CM signal sequence, and a CJ signal sequence; G 1 is shown sequentially providing an ANSam signal sequence and a JM signal sequence; G 2 is shown sequentially providing a CNG signal sequence, a CM signal sequence, and a CJ signal sequence; and F 2 is shown sequentially providing an ANSam signal sequence and a JM signal sequence. The design and operation of these signal sequences are understood in the art. The CNG, CM, CJ, and JM signals are defined and described in the ITU V.8 Recommendation, which is incorporated herein by reference as if set forth in its entirety.

In addition to the standard T.38 signals, the signalling between the calling or emitting side and the answering or receiving side includes signals to accomplish network latency recovery. The standard T.38 signals shown between the gateways G 1 and G 2 are t30-data(JM, info), t30-data(CM, info), and t30-indicator(CNG). The T.38 signals added for network latency recovery are a t30-indicator(CNG sent) signal and a t30-indicator(DoV.8/Dolowspeed) signal. Certain signals are designated “t30” to reflect that the terminals F 1 and F 2 communicate during a T.30 session. The nature of a T.30 session is understood in the art. The designation also relates to the type of message in the particular signal. The t30-indicator (DoV.8/Dolowspeed) signal is shown as the t30-indicator(DoV.8) signal in FIG. 1 and is shown as the t30-indicator(Dolowspeed) signal in FIG. 3 . The t30-indicator(CNG sent) signal is used to indicate to G 1 if a CNG response has been transmitted by G 2 . The t30-indicator(DoV.8/Dolowspeed) signal is used to indicate whether a V.8 modem connection should continue or if a Group 3 modem connection should be established. Network latency is here defined as the time between transmission of the t30-indicator(CNG) signal and reception of the t30-indication(CNG sent) signal. If network latency is acceptable, then G 1 responds with the t30-indicator(DoV.8) signal ( FIG. 1 ). If network latency is unacceptable, then G 1 responds with the t30-indicator(Dolowspeed) signal ( FIG. 3 ). Normally, G 2 and F 2 have up to 4 seconds to respond with a V.8 signal, so it is proper to allow G 2 to continue with CNG until G 2 receives the t30-indicator(DoV.8/Dolowspeed) signal.

Referring to FIGS. 2A–2B , a flowchart of an exemplary network latency recovery process in accordance with the signalling and communications system C of claim 1 is shown. The process begins in step 200 where F 1 transmits a CNG to G 1 . Next, in step 202 , G 1 detects the CNG and in response sends the t30-indicator(CNG) signal to G 2 . G 2 is thereby informed to start transmitting CNG. From step 202 , control proceeds to step 204 where G 1 starts its timer 400 ( FIG. 4 ). Control next proceeds to step 206 where it is determined by G 1 if a CNG sent signal has been received from G 2 . Control remains in step 206 until the CNG sent signal is received from G 2 . When the CNG sent signal is received, control proceeds to step 208 where G 1 stops its timer 400 . Based on the starting time and stopping time of the timer 400 , G 1 now stores a network latency timer value 408 ( FIG. 4 ). When the network latency timer value 408 is too long, the high speed modem connection (in this case, a V.8 connection) fails. Until the CNG sent signal is received, G 1 holds off transmission of an ANSam signal.

Next, in step 210 , it is determined if the timer value 408 is below an acceptable latency threshold 404 ( FIG. 4 ). If so, then control passes to step 212 where G 1 transmits the t30-indicator(DoV.8) signal to G 2 . Next, in step 213 , G 1 continues transmitting ANSam since the timer value 408 was below an acceptable latency threshold 404 . From step 213 , control proceeds to step 214 where G 2 ceases CNG transmission and transmits CM based upon the CM contents of F 1 . Next, the network latency recovery process is completed in step 216 .

Referring to FIG. 3 , the communications system C and associated signalling is shown for an unacceptable network latency. The F 1 is shown sequentially providing two CNG signal sequences and a DCS signal sequence; G 1 is shown sequentially providing a CED signal sequence and a DIS sequence; G 2 is shown sequentially providing two CNG signal sequences and a DCS signal sequence; and F 2 is shown sequentially providing an ANSam signal sequence and a DIS signal sequence. The design and operation of these signal sequences are understood in the art. In terms of FIG. 2 , steps 210 – 214 are specific to FIG. 3 . Steps 200 – 10 are applicable to both FIG. 1 and FIG. 3 . In step 210 , if the timer value 408 is above the acceptable latency threshold 404 , then control passes to step 218 where G 1 transmits the t30-indicator(Dolowspeed) signal to G 2 ( FIG. 3 ). Next, in step 220 , G 1 transmits CED, instead of ANSam, forcing F 1 to use V.17 (a standard for data signalling rates up to 14,400 bit/s) and lower procedures. A switch from one modem version to a previous modem version has generally been termed a “fallback.” G 1 continues to send CED while the timer value 408 is above the acceptable latency threshold 404 . From step 220 , control passes to step 222 where G 2 continues with its CNG transmission, forcing F 2 to use V.17 and lower procedures. The network latency recovery process is then completed in step 216 .

Referring to FIG. 4 , exemplary components of a calling modem M in the calling gateway G 1 is shown. The modem M includes a comparison block 402 coupled to a timer 400 , a latency threshold 404 , and a modem connection (or modulation mode) selection block 406 . The timer 400 stores the network latency value 408 described above. The comparison block 402 compares the network latency value 408 to the network latency threshold 404 described above. Based on the result of that comparison, the modem connection selection block 406 generates either the t30-indicator(DoV.8) signal or the t30-indicator(Dolowspeed) signal. If the network latency value 408 is less than the latency threshold 404 , then the modem connection selection block 406 generates a t30-indicator(DoV.8) signal. If the network latency value 408 is greater than the latency threshold 404 , then the modem connection selection block 406 generates the t30-indicator(Dolowspeed) signal. It should be understood that the illustrated components can be implemented in a variety of ways (hardware or software).

Thus, when network latency is greater than an acceptable value, a high speed modem connection is bypassed for a low speed modem connection. The high speed modem connection, for example, can be a V.8 modem connection, and the low speed modem connection can be a Group 3 modem connection. Also, the high speed modem connection can be a V.34 half-duplex connection. It should be understood that the network latency recovery techniques described herein are applicable to a variety of modem connections or modulation modes. In addition, while network latency recovery is described in a modem context, it should be understood that the network latency recovery techniques described herein are applicable to other communication devices with similar structure and/or operation. Other ways of suspending or terminating a high speed modem connection based on current network latency without departing from the spirit of the invention are possible. Likewise, other ways of selectively operating a high speed modem connection and a low speed modem connection based on current network latency without departing from the spirit of the invention are possible.