MULTIPLE PROTOCOL COMMUNICATIONS

Abstract

A particular method includes receiving a dialing code from a calling device at a customer premise device coupled to the calling device. The method also includes selecting a telecommunication protocol based at least in part on a network of a receiving device. The method further includes initiating a call to the receiving device using the selected telecommunication protocol.

Description

FIELD OF THE DISCLOSURE

The present disclosure relates generally to implementation of communication protocols.

BACKGROUND

Less than thirty years ago, the term “telecommunications” connoted making and receiving telephone calls over the public switched telephone network (PSTN). Today, telecommunications means transporting data representing voice, video, text, and instructions over wired and wireless digital networks such as the Internet.

Within the PSTN telephony environment the equipment needed to support the telecommunications infrastructure was centralized at a telephone company “central office” so that Customer Premises Equipment (CPE) could be limited to simple telephones. The nature of modern digital networks is to decentralize many functions and capabilities thus requiring more complex CPE to provide access. However, subscribers expect newer digital telecommunications to be usable with the same ease as the traditional telephone and at low cost. This expectation dictates that the digital network interfaces and associated protocols be compact and unobtrusive, implemented inexpensively, and require little in the way of subscriber interaction.

The complexity of implementing a CPE for digital telecommunications comes primarily from the need to implement the protocols used to organize information sent over digital networks. These protocols evolved in rich computing environments with many computing resources (e.g., central processing unit (CPU) power for computation; memory for data and program storage). Additionally, the protocols evolved quickly (i.e., in months or years, vs. the traditional PSTN years or decades), reflecting knowledge gained from actual application. The CPE for digital communications of today must, therefore, have an effective way to deal with protocol implementation within the restricted computational resources dictated by the CPE's restricted hardware cost.

What would be useful is a system and method for implementing multiple digital telecommunication protocols on a reduced hardware CPE that is not limited to any specific protocol.

SUMMARY

A particular embodiment is a telecommunications gateway that implements telecommunication protocols using a telecommunication protocol engine (TPE). Telecommunication protocols include multiple digital networking protocols (e.g., Session Initiation Protocol (SIP), H.323, DHCP, TCP/IP and STUN protocol) and telephony protocols. However, the present disclosure is not so limited. As will be apparent to those skilled in the art, any protocol that facilitates telecommunications over digital networks (both between digital devices, a digital device and an analog device, and between analog devices) may be implemented by the TPE without departing from the scope of the present disclosure.

In this embodiment, the TPE is implemented using inexpensive, memory limited microprocessors and inexpensive flash memory. However, this is not meant as a limitation. As will be apparent to those skilled in the art, embodiments may be implemented in other computing contexts without departing from the scope of the present disclosure.

Therefore, an aspect of the present disclosure is an implementation of a telecommunication protocol engine (TPE) using a memory limited microprocessor and flash memory.

Another aspect of the present disclosure is an implementation of a (TPE) using a “virtual machine” that executes instructions from flash memory.

Another aspect of the present disclosure is the representation of telecommunication protocols as Finite State Machine (FSM) abstractions.

Still another aspect of the present disclosure is the implementation of a virtual machine to support FSMs used to express protocol implementation.

A further aspect of the present disclosure is the implementation of FSMs using virtual machine instructions stored in a flash memory, wherein the virtual machine instructions represent telecommunication protocols implemented as states, instructions, and transitions.

Yet another aspect of the present disclosure is a CPE Control Protocol that specifies how an end user can control the behavior of the CPE using a standard telephone that may be connected directly to the CPE or that accesses the CPE remotely.

An aspect of the present disclosure is a telephony gateway including a TPE.

Another aspect of the present disclosure is the implementation of the Session Initiation Protocol (SIP), H.323 protocol, DHCP and STUN protocol as FSM.

These and other aspects of the present disclosure will become apparent from a review of the general and detailed descriptions that follow. A particular embodiment is a telecommunications gateway that implements multiple digital networking protocols using a telecommunication protocol engine (TPE). In this embodiment, the TPE is implemented using inexpensive, memory limited microprocessors and inexpensive flash memory. However, this is not meant as a limitation. As will be apparent to those skilled in the art, particular embodiments may be implemented in other computing contexts without departing from the scope of the present disclosure.

A finite state machine (FSM) execution facility is implemented using virtual machine instructions located in the flash memory. The “state” of a given instance of a FSM is located in the RAM and accessed by the microprocessor. As the virtual machine executes instructions from the flash memory, it modifies the FSM state in RAM along with accessing other facilities of the microprocessor and other software resources.

Another embodiment includes a method for representing a protocol as a FSM using an extension of the C++ programming language. Specifically, an FSM specification may be created on a development computer using C++ with a specialized library. The result is a program that, when executed on the development computer, produces virtual machine instructions that can be loaded into the flash memory of the TPE. The virtual machine within the TPE microprocessor then uses these instructions as described above. While this embodiment uses the C++ programming language and a C++ library, the present disclosure is not so limited. As will be appreciated by those skilled in the art, other programming languages (and related libraries) may be used to produce virtual machine instructions that define an FSM.

An additional aspect of the present disclosure is the specification of a CPE Control Protocol that is implemented using the technique described above. This protocol specifies how an end user can control the behavior of the CPE using a standard telephone that may be connected directly to the CPE or accessing the CPE remotely by “calling” over either a Voice over Internet Protocol (VoIP) or PSTN connection. This protocol allows the user to direct the CPE to place a local telephone to VoIP call; a local telephone to local PSTN call or a received VoIP call routed to the local PSTN based call. Additionally, this protocol allows the user to modify other operations of the CPE.

The CPE Control Protocol receives input from the user via the standard telephone touch-tone keypad. Specifically, the user enters a pound-sign (#), a sequence of digits or stars identifying the operation with any related data and a terminating pound-sign (#) indicating the end of user input. The CPE Control Protocol communicates with the user via one or more facilities depending on the originating location of the command. These include: flashing of light-emitting diodes (LEDs) on the CPE; generation of tones played over the telephone; voice commands played over the telephone; placing a call back to the telephone; and using a “Caller ID” mechanism to present alpha numeric data via the telephones Caller ID display facility.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1A illustrates a functional block diagram of a telecommunication protocol engine according to an embodiment.

FIG. 1B illustrates a functional block diagram of a telecommunication protocol according to an embodiment.

FIG. 1C illustrates a function block diagram showing data flows in a telecommunication protocol engine according to an embodiment.

FIG. 2 illustrates a flowchart of a process for implementing a telecommunication according to an embodiment.

FIG. 3 illustrates means for expressing a finite state machine (FSM) according to an embodiment.

FIG. 4 illustrates a Voice over Internet Protocol (VoIP) telecommunication between a caller and a called party from the perspective of the caller according to an embodiment.

FIG. 5 illustrates a script created using a scripting language to implement the protocol illustrated in FIG. 4.

FIG. 6 illustrates a block diagram of a telephony gateway according to an embodiment.

FIG. 7 illustrates a service offering according to an embodiment.

FIG. 8 illustrates a service provider gateway according to an embodiment.

DETAILED DESCRIPTION

A particular embodiment is a telecommunications gateway that implements multiple protocols using a telecommunication protocol engine (TPE). In this embodiment, the TPE is implemented using an inexpensive, memory limited microprocessor and inexpensive flash memory. A finite state machine (FSM) execution facility is implemented in firmware using virtual machine instructions located in the flash memory. The “state” of a given instance of a FSM is located in the RAM and accessed by the microprocessor. As the virtual machine executes instructions from the flash memory, it modifies the FSM state in RAM along with accessing other facilities of the microprocessor and other software resources. However, this is not meant as a limitation. As will be apparent to those skilled in the art, particular embodiments may be implemented in other computing contexts without departing from the scope of the present disclosure.

Referring to FIG. 1A, a block diagram of a TPE 100 according to an embodiment is illustrated. The TPE 100 includes microprocessor 102 and flash memory 140. The microprocessor 102 includes a central processing unit (CPU) 120, random-access memory (RAM) 115, and firmware 105 in which a virtual machine 125 resides. The CPU 120 runs the virtual machine 125 and communicates with the flash memory 140 over a data bus 110. In the preferred embodiment, the data bus 110 is a parallel bus, but this is not meant as a limitation. In another embodiment, the data bus 110 is a serial bus. The flash memory 140 holds protocol templates for protocol A 145, protocol B 155, and protocol N 160 and data relating to each protocol. Each protocol template includes virtual machine instructions 150 defining one or more finite state machines (FSMs) implementing a particular protocol, illustrated only for protocol A 145. The RAM 115 may include finite state machine (FSM) instructions.

A protocol template provides specifications for the one or more FSMs that represent a single protocol. Each FSM specification is used to generate a set of virtual machine instructions 150 that define a FSM implementing all or a portion of a particular protocol.

FIG. 1B illustrates a functional block diagram of a telecommunication protocol according to an embodiment. Protocol A 145 includes the virtual machine instructions 150. The virtual machine instructions 150 for protocol A 145 define a first finite state machine 161, a second finite state machine 162, and an Nth finite state machine 163. Thus, a single protocol may include more than one finite state machine. Telecommunication protocols include multiple digital networking protocols (e.g., Session Initiation Protocol (SIP), H.323, dynamic host configuration protocol (DHCP), transmission control protocol/internet protocol (TCP/IP) and session traversal utilities for network address translation (STUN) protocol) and telephony protocols. However, the present disclosure is not so limited. As will be apparent to those skilled in the art, any protocol that facilitates telecommunications over digital networks (both between digital devices, a digital device and an analog device, and between analog devices) may be implemented by the TPE without departing from the scope of the present disclosure.

The virtual machine instructions 150 are read and executed on demand by a virtual machine that resides in the firmware 105. Since the microprocessor 102 (see FIGS. 1A and 1C) retrieves instructions in response to a request from the virtual machine 125, only a tiny portion of an entire protocol specification (specifically one virtual machine instruction and its associated operands) is stored in the RAM 115 at any given time. One advantage of this arrangement is that the RAM 115 contains only the currently executing virtual machine instruction and data, and a few miscellaneous data structures needed for representing the current FSM states (also fixed in size). The content of the flash memory 140—which can be quite sizeable if it implements a number of complex protocols—does not utilize any significant portion of the RAM 115. The size of the virtual machine instructions 150 is limited by the amount of flash memory 140 available, and has no significant impact on the RAM 115 resources. This allows complex protocols to be implemented on very low cost microcontroller architectures in which program and data memory is very limited but flash memory is plentiful.

FIG. 1C illustrates a function block diagram showing data flows in the telecommunication protocol engine 100 according to an embodiment. The bus references have been deleted for clarity. A protocol creation system 175 produces the virtual machine instructions 150. In an embodiment, the protocol creation system 175 includes a C++ development environment that is used to encode an FSM specification for a protocol as human readable text. The resulting C++ program is then executed to generate the virtual machine code that is store in the flash memory 140 via I/O ports 165 and I/O bus 170 (FIG. 1A). While this embodiment uses the C++ programming language and a C++ library, the present disclosure is not so limited. As will be appreciated by those skilled in the art, other programming languages (and related libraries) may be used to produce virtual machine instructions that define a FSM.

A virtual machine instruction 150 and its operands are retrieved, by the CPU 120 at the direction of the virtual machine 125 and stored in RAM 115 for execution. Virtual machine 125 then executes the instruction that was retrieved.

A telecommunication protocol is implemented in accordance with the process embodiment illustrated in FIG. 2. A call is made to the TPE to implement a telecommunication protocol, at 200. The call is processed, at 205, by the microprocessor via the TPE software, which determines, at 210, which telecommunication protocol is to be implemented and where the protocol template resides in flash memory, at 215. FSM data is then created in RAM and initialized to represent the state of the FSM used to implement the protocol, at 220. The virtual machine then starts executing instructions from a location within the flash memory based upon the FSM data, at 225.

The classic definition of a FSM is a collection of states. When a FSM is being executed, it is said to be “in” a specific state waiting for an event. When an event occurs, the FSM definition asserts a next state to be entered. When a state is entered a set of actions may performed, then the FSM waits in that state for the next event. This FSM operational model is implemented directly by the virtual machine described above in reference to FIG. 1B and supported by the specialized virtual machine instructions generated from the FSM that expresses the protocol. Specifically, the State entry actions are expressed as specialized virtual machine instructions. Additional the virtual machine is capable of executing multiple FSM at the same time be they multiple instances of one FSM definition or different FSM definitions. It is through this facility that the TPE supports multiple protocols concurrently.

The virtual machine in this embodiment is specially designed to operate with events modeled as “tokens” so that it can respond to both physical events identified by the microprocessor firmware or logical events generated by other FSMs being concurrently executed in a uniform manner. The representation of “tokens” and their management further enhances the TPE to operate on very low cost microprocessor architectures in which program and data memory is very limited.

Virtual machine instructions are generated using a “translator” that receives human-readable syntax and translates this syntax to FSM instructions. To facilitate the specification of a protocol in human readable form and support its maintenance as the protocol evolves over time, the translator and virtual machine support the concept of shared state entry instructions through function or macros. These are collections of virtual machine instructions that reside in the flash memory, along with FSM specifications. As required, the virtual machine can execute these special collections of instructions upon demand.

FIG. 3 illustrates a means for expressing FSM according to an embodiment. Referring to FIG. 3, a function name is associated with a purpose. These function name/purpose pairs are used to author protocol implementations via FSM definitions. A function name/purpose pair (referred to as an “Advocate”) corresponds to a facility “known” to the virtual machine that is implemented in the firmware. While this exemplary embodiment uses the C++ programming language, the present disclosure is not so limited. As will be appreciated by those skilled in the art, other programming languages may be used to produce virtual machine instructions that define a FSM.

The C++ definition of each Advocate includes code that produces the appropriate virtual machine instructions that will be placed into the flash memory. Note that some Advocates take parameters. The task of setting up virtual machine instructions for accessing and updating these parameters is also performed by C++ code within the body of the Advocate.

FIG. 4 illustrates a VoIP telecommunication protocol according to an embodiment between a caller and a called party from the perspective of the caller. FIG. 5 illustrates a script according to an embodiment created using the scripting language (FIG. 3) to implement the protocol illustrated in FIG. 4. The caller initiates a call, at 400, causing the FSM to send a message to the called party, at 410. Referring to FIG. 5, the initial state of the FSM is “CALL INITIATED.” The FSM then receives signaling messages via the network from the called party indicating the response (status) of the called party, at 420. If the called party is busy, the FSM receives a CALLED PARTY BUSY condition (referred to as an event), at 430, executes a PLAY BUSY TONE command (an action), at 460, and enters a “BUSY TONE” state, at 475. Similarly, if there is no response from the called party and the call times out, at 420, the FSM receives a TIME OUT condition, at 440, executes a PLAY FAST BUSY TONE command, at 470, and enters a “FAST BUSY TONE” state, at 485. If the call is delivered, the FSM receives a CALL DELIVERED condition, at 450, executes both an “Init Vocoder” and “Send (CONNECT_ACK)” instruction, at 465, and enters a “Voice” state, at 480.

FIG. 6 illustrates a block diagram of a telephony gateway 600 according to an embodiment. A digital signal processor (DSP) 603 communicates with a CPU 610 over a host interface/API 605. Power to the telephony gateway 600 is provided via a power supply 615. The DSP 603 communicates with a subscriber line interface (SLI) 620 that offers connection to consumer premises equipment through a first external jack 625. The SLI 620 interfaces with customer premises equipment, such as a telephone. The DSP 603 communicates with a data access arrangement (DAA) 635 that offers connection to a public switched telephone network (PSTN) through a second external jack 637.

The CPU 610 is connected to a RAM 650, to a flash memory 640, and to a firmware unit 660. Together, the CPU 610, the RAM 650, the flash memory 640, and the firmware unit 660 include a telecommunication protocol engine as described in the context of FIG. 1. The CPU 610 also provides an interface to I/O ports 630. In an embodiment, the I/O ports 630 include a serial interface and an Ethernet interface. The I/O ports 630 are accessible through a third external jack 645. It is anticipated that as RAM becomes cheaper and more capable of enhanced storage that the firmware unit 660 can also reside within the RAM 650.

FIG. 7 illustrates a service offering according to an embodiment. Referring to FIG. 7, a service provider network 700 includes a service provider proxy 703, a service provider gateway 705 and connectivity to a public switched telephone network (PSTN) 740, an internet service provider (ISP)-A network 710, and an ISP-B network 715. A telephony gateway-A 720 is in communication with the ISP-A network 710. A telephone-A 725 is connected to the telephony gateway-A 720. Similarly, a telephony gateway-B 730 is in communication with the ISP-B network 715. A telephone-B 735 is connected to the telephony gateway-B 730.

A connection between the telephony gateway (720,730) and its associated ISP network (710, 715) is made by means known in the art. By way of illustration and not as a limitation, the telephony gateway-A 720 is in communication with the ISP-A network 710 using DSL connection. The telephony gateway-B 730 is in communication with the ISP-B network 715 via a dialup connection. It is noteworthy that no general-purpose computer is required to establish communication between the telephony gateway (720, 730) and the service provider gateway 705.

Both the telephony gateway-A 720 and the telephony gateway-B 730 are registered with the service provider network 700. The telephony gateway-A 720 may place a call to the telephony gateway-B 730 by interacting with the service provider network 700. The service provider proxy 703 then contacts the telephony gateway-B 730 on behalf of the telephony gateway-A 720 to facilitate call setup (also known in the art as signaling). Once the signaling has been completed, the telephony gateway-A 720 interacts directly with the telephony gateway-B 730 passing information until the call is terminated. When the call is terminated, call tear down signaling is performed through the service provider proxy 703.

Both the telephony gateway-A 720 and the telephony gateway-B 730 are registered with service provider network 700. The service provider network 700 is also connected to the PSTN 740, which is connected to a telephone-C 745 and a telephone-D 750. A call placed by telephony gateway-A 720 to the telephone-C 745 via the PSTN 740 would involve call setup and tear down signaling between the telephony gateway-A 720, the service provider proxy 703, and the service provider gateway 705.

A call placed over the service provider network 700 is routed by the service provider gateway 705. In an embodiment, the protocol used by a telephony gateway (720, 730) and the routing is controlled by a number dialed to initiate a telephone call via a CPE Control protocol. By way of illustration and not as a limitation, a call placed from the telephone-A 725 to the telephone-B 735 is an “on-network” call, meaning both the calling party and receiving party are using registered telephony gateways (720, 730). In this embodiment, the telephone numbers associated with registered telephony gateways begin with the same prefix, for example 777. In this embodiment, the calling party using the telephone-A 725 presses #777 (plus the remaining telephone number digits)# on the telephone-A 725 (note that the starting and ending pound-sings (#) reflect the requirements of the CPE Control protocol). The service provider gateway 705 determines from the prefix preceding the remaining telephone number digits that the call is on-network and connects the telephone-A 725 to the telephone-B 735 over the service provider network 700.

By contrast, if the telephone-A 725 places a call to the telephone-C 745, the call is dialed using a standard prefix (1+areacode+number) again bracketed by the pound-signs required by the CPE Control protocol (e.g., #16503288459#).

As will be apparent to those skilled in the art, other signaling conventions may be used to route telephone calls without departing from the scope of the present disclosure.

In an embodiment, a service provider gateway is a self-contained system for providing telephone communications using telephony gateways as embodied herein. FIG. 8 illustrates a service provider gateway according to an embodiment.

Referring to FIG. 8, a service provider gateway 800 includes network communication systems 805 and protocol handling systems 815. These systems permit the service provider gateway 800 to communicate with various networks using protocols required by the networks or required by the type of communication service being provided. Interactive voice response (IVR) software 830, in combination with a billing system 835 and data management and storage systems 850 gather billing data and automate billing questions. Authentication of users is managed by an authentication system 825.

In yet another embodiment, a service provider gateway 800 according an embodiment as described in reference to FIG. 8 is licensed for use by a third party service provider (TPSP). In one embodiment, the TPSP is permitted under the license to provide service to subscribers using a telephony gateway as described previously wherein the service is private labeled in the name of the TPSP. In an alternate embodiment, the TPSP acquires franchise rights to provide services to subscribers in the name of the licensor. In still another embodiment, the TPSP is permitted to offer access to a service provider network operated by the licensor in the name of the TPSP.

In yet another embodiment, the telephony gateway is configured to receive communications from a remote location via a telephone call (either from the PSTN or a wireless service provider). The communications may be used to configure the gateway or to initiate a call from the gateway to a remote communication device. In this latter embodiment, the gateway additionally functions as a bridge between the incoming calling device and the remote communication device. The gateway answers the incoming calling, the CPE Control protocol accepts user input (e.g., an authentication personal identification number (PIN), star and target phone number: #65432*7771234567#) necessary to call the remote receiving device, and then places a VoIP call to the remote communication device.

In another embodiment, two telephony gateways are connected to first and second communication devices respectively that are in communication with each other. The first communication device sends a “hook-flash” signal to the first telephony gateway. The first telephony gateway suspends the communication with the second telephony gateway and enables caller access to the CPE Control protocol. Using this protocol the user directs the CPE to initiate a three-way call to another phone. The CPE Control protocol will notify the CPE Control protocol on the second device that a three-way call has been initiated. The three-way connection includes sharing data from one party with the other two parties on the call. In this manner, a three-way connection is established and maintained without external mixing devices or the need to deploy a media gateway in the VoIP system.

The traditional VOIP architecture contained three layers: A Signaling Layer, which converted signaling information between PSTN protocols (such as CAS, SS7 or ISDN) and VOIP protocols (such as MGCP, Megaco, or R323); a Switching Layer that routed and switched VOIP calls to Signaling, Switching (for other systems) or the Application layers; and an Application Layer, which provided IP-centric enhanced voice services.

Prior to adopting H.323, there were few if any interoperability standards. First-generation gateways were closed, proprietary systems unique to each manufacturer. Few worked with one another. While well suited to toll bypass applications, their proprietary design severely limited their ability to interconnect to carrier grade networks. Most could not support voice services beyond toll bypass and debit card applications.

The subsequent generation VOIP call control architecture used H.323 protocol to accomplish simple tasks like call setup/control between gateways or billing/rating and routing functions for multipoint networks. Call control software acted simply as middleware that directed the media gateways as to how to set up and tear down calls between gateways. Operating in real time, the call control software kept track of which media gateways were available, where they were located, and often determined how to route calls over the network. Soft-switches evolved out of this climate to overcome the limitations of the Media Gateway Control Protocol (MGCP) and the media gateway controllers to support voice services applications. Soft-switches are designed to support PSTN/IP internetworking and signaling protocol mediation, (e.g., out-of-band SS7/C7, in-band R1, CAS, etc.) to switch calls and control the media gateways in its domain through MGCP or Megaco, and in addition, support application interfaces such as the Session Initiation Protocol (SIP). This enabled a system to interact in real-time with voice streams. Further, these enhancements allowed a system to respond in real-time to detect dual-tone multi-frequency (DTMF) or voice-recognition inputs and responses.

But these systems are incapable of handling DSP-intensive media processing, for example, such as the type of processing required to handle Interactive Voice Response (IVR) and conferencing applications. Media Servers were introduced to handle these tasks, but they created Quality of Service (QoS) problems. This is because, streaming VOIP placed an entire application at the mercy of the QoS level that an underlying IP network could provide. Dropped packets because of instability of the underlying IP network could negatively affect the overall QoS. Moreover, streaming cannot support the basic, everyday telephony functions such as unified messaging, store-and-forward faxing and other standard functions that are needed today. This requires DSP-intensive media processing, especially for in-band DTMF and fax tones. Further, streaming solution could tax the system resources because the system could be required to process a large number of small packets, each containing the normal TCP/IP overhead.

In an illustrative architecture of a VOIP-enabled telecommunication system, a calling party device—first Customer Premise Equipment (CPE)—is coupled to a network, such as a packet switched network (PSN), such as the Internet. A called party device—second CPE—is coupled to the PSN.

Microgateway—a VOIP-Enabled CPE

In an aspect, the instant disclosure is directed toward a CPE called a microgateway (MG), which connects telecommunication terminal equipment such as a telephone handset to a PSN. Microgateway could be connected to any other terminal device, such as a desktop or a handheld personal computer, or a telephone handset. Alternatively, a general-purpose server computer could be coupled to the PSN, which computer is programmed suitably to provide the functions of a microgateway (MG). Note that the microgateway is not the same in form or function as a Media Gateway or a VOIP gateway. Rather, microgateway is the name given to a CPE device connected to a terminal device and configured to work as a single port gateway, as will be explained in detail in the following.

In an embodiment, microgateway includes a processor, some memory and communication devices to couple the microgateway to a network and to receive and transmit information in digital or analog form. Additionally, where the terminal device coupled to the microgateway is a telephone handset—which could be a digital or analog telephone handset—the microgateway could connect to the PSN via a telecommunication switch such as Lucent 5ESS® or equivalent local office switch. Where the terminal device is a mobile handset, it could be coupled to the PSN via a wireless telecommunication network, and the details of how these connections are established are known to persons of ordinary skill in the art and need not be repeated here. The microgateway is configured to interface with both Ethernet and dial-up VOIP lines. It is understood that the first and the second CPEs (microgateways) are connected to the PSN with an address such as a dotted-decimal address used in IP addressing, for example, 121.255.48.16, which is known to persons skilled in the art.

The microgateway could be implemented in the form of custom hardware or in software. In a hardware implementation, the microgateway includes custom-manufactured Application-Specific Integrated Circuits (ASIC) that could be programmed with application logic, and other supporting circuit elements, such as digital signal processors (DSP) and memory devices to process signals. These details will be explained with reference to microgateway hardware architecture below.

In an embodiment, a block architectural diagram of the microgateway contains a Digital Signal Processor (DSP), such as a VP 1405; a Central Processor Unit (CPU), such as an IP2022 Network Processor; memory, such as semiconductor memory; devices to interface with a variety of input/output (I/O) or peripheral devices, such as Ethernet, Serial I/O interface, Subscriber Line Interface Card (SLIC), Data Access Arrangement (DAA), and the like; and a suite of software programs, such as a Telephony Protocol Engine and protocol stacks, such as H.323, SIP (for VOIP), TCP/IP, Real Time Protocol (RTP), and V.34 soft modem programs. It should be noted that the microgateway may be implemented totally in software, where the critical DSP functions could be implemented as software code executed in a general purpose processor such as a Pentium III® microprocessor. It is assumed that the first and the second CPEs are connected to the PSN either directly, via a modem, or via another network, which other network could be a mobile or PSTN or the like. The principles of the disclosure apply equally well for all these variations.

A microgateway may be integrated into a VOIP architecture in various forms (either in the form of a circuit board or in the form of an AMC chip). For example, a microgateway may be coupled to a traditional telephone and thereafter to both a PSN and the PSTN via different interfaces, thereby making a “regular” PSTN telephone call as well as a VOIP call (via the PSN) possible. In another example, a low-cost dialup adapter may be coupled to the microgateway enabling a connection to an ISP and a VOIP telecommunication network. In another example, the microgateway could be used to interface with a variety of telephony I/O devices, such as an analog/digital wireline telephone handset, a cordless telephone handset, and, via optional I/O ports, to a keypad such as either a QWERTY keyboard or a DTMF keypad, and an LCD. In another example, the microgateway (e.g., an ASIC chip) could be incorporated into numerous devices.

A software architecture of the disclosed system includes a transport layer such as the Transmission Control Protocol (TCP) or the User Datagram Protocol (UDP) layer over an Internet Protocol (IP) layer. It should be noted that this description provides only an embodiment of the principles of the disclosed system and therefore the principles should not be so limited. The IP layer may be configured to function over an Ethernet (the protocol of choice for local area networks) or a PPP interface (which is the protocol of choice for a dialup ISP connection) layer. A network layer (including the layers for Signaling, Call Control and Media control) resides above the transport layer. A Media Gateway Application Program Interface (Mg API) layer is configured above the network layer. Application layer and the Telephony Protocol Engine (TPE) layers reside above the network layer. An API layer for customized user applications and DSP-intensive VOIP applications (such as conference calling, three-way calling, etc) resides above the Application layer. Standard telephony protocols such as the G.168 (Echo Canceler), G.711, G.723 (Voice Codec), G.729 and G.726 are also programmed within the microgateway. And finally, codec drivers and interface software or firmware for SLIC, DAA and other interfaces is also provided within the microgateway.

Virtual Memory Subsystem

Traditional designs for telephony applications used multiple processors and a shared main memory that stored data. Sometimes, each processor may use a local cache to store its own private copy of a subset of the data in the main memory. In this traditional system, a separate memory control chip connecting the processors to the main memory manages the operations necessary to access memory from the processor. For example, if one of the processors makes a reference to memory to access data, the control chip will schedule a read for the data corresponding to the reference from the main memory while simultaneously scheduling system probes to the other processors to check for the presence of this address in the processor cache.

In general, to establish a connection between the microprocessor and the memory, the control chip uses 2*k pins with k bits for each of two address buses, i.e., the address-out bus for communicating the address of the memory reference from the processor to the control chip, and the address-in bus used by the control chip to send an address of a probe into each of the processors. If k is large, the control chip can quickly become pin-limited and cannot address a large amount of memory.

It is desirable to manage or address a large amount of memory without the limitations imposed by low pin counts. Though some large pin count processors exist, they are expensive. And low-cost telephony applications, such as the disclosed VOIP application, would benefit from a low pin count architecture, while achieving a larger addressable memory via a virtual memory system. It is generally known that a processor with a single address bus for all memory size requirements will still require all the address pins be connected to the control chip even though a lesser number of pins are actually needed.

Thus, it is desired to reduce the pin count of the memory control chip for a multiprocessor system with a limited memory size requirement by providing a microprocessor which permits a designer to select an address bus supporting one of a maximum memory size requirement and a small memory size requirement.

In general, pin count limitation increases the number of chips needed to emulate a particular logic design and thereby decreases emulation speed, because signals must cross more chip boundaries, and increases system cost. This is especially pronounced in microcontroller chip designs where reducing the overall cost is a critical factor.

The instant disclosure includes a method of overcoming pin count limitations using a virtual memory system that uses a virtual addressing scheme. VMS is built on a DSP/micro-controller and allows in conjunction with the TPE to run high-level protocols, such as H.323, which requires the ability to address a large amount of memory.

Telephony Protocol Engine

The microgateway includes a Telephony Protocol Engine (TPE). This TPE enables designing a system with protocol-independent architecture, thereby enabling one to support a variety of protocols, such as H.323. Additionally, using the TPE, one can execute complex protocols on memory-limited and I/O (pin-limited) micro-controllers or DSP chips.

The process of off-loading the state transitions, semantic rules and language tokens to an external flash memory is called uploading a template. In an aspect, rules of a protocol are abstracted and stored on an auxiliary memory, which can be accessed using a parallel bus or serial link that connects the auxiliary memory to the TPE.

A protocol-independent machine can be built by adopting a generalized state machine that can determine its next state from the current state and an input. The protocol engine could use generic tokens that represent each state, thereby enabling the implementation of a variety of protocols with little complicated coding. Additionally, the protocol engine could be stored in an auxiliary memory, thereby enabling a designer to use the virtual memory features to address a larger address space than is permitted by the traditional pin-count limitations of a moderately priced processor. Moreover, the auxiliary memory could be a volatile memory (in which case the protocol engine should be loaded at power up) or a non volatile memory (in which case the protocol engine could be permanently burned into an EPROM or a Flash memory for easy reuse).

In an embodiment, the architecture for the TPE and the VMS both may be implemented entirely in hardware or in software. In this embodiment, state information, semantic rules, tokens and data tables that are necessary to execute a telephony protocol are stored in an auxiliary memory device, such as the Atmel AT45DB011 flash memory. An aspect of the implementation is that the TPE operates in conjunction with the main CPU, which executes program instructions that act upon the state information, semantic rules, rules and data tables to effect state transitions.

Common Channel Signaling to Setup and Manage VOIP Calls

Elements in the disclosed system, including the microgateway, may also be configured to transport voice packets over the SS7/C7 protocol. The SS7/C7 protocol allows the switch to use out-of-band signaling. Out-of-band signaling refers to the capability of the protocol to send signaling information separately from the voice circuits. This greatly increases call processing and trunk efficiency in the disclosed system as speed for a call set up can be increased. Moreover, the disclosed system has the ability to not use trunks if the called party phone is busy. The trunks that would have been used to make the call can then be used to make other calls. Further, the system designed herein provides the ability to release calls to the previous gateways if the call fails for a user-termination reason other than user-busy.

Remote Access and Control of Microgateway Via a Mobile Telephone Handset

The disclosed microgateway is additionally configured to receive communication from a remote location via a telephone call. Such information may be used to configure, control and maintain the microgateway, or to initiate additional calls from the microgateway.

As stated above, the microgateway is equipped with a SLIC and LAN or dialup VOIP interfaces, in addition to other interfaces. A user operating a regular telephone handset may make a telephone call into the microgateway by dialing a number designated to the microgateway. In this case, the microgateway functions as a regular telephone destination CPE, but the call terminates with the microgateway. The user can then enter the required data to the microgateway in order to initiate an outbound call from the microgateway over the LAN or other VOIP connection. Advantageously, the microgateway establishes a bridge connection to a second destination, thereby allowing the user to communicate with a party connected to the second destination. This is helpful in cases where a user is out of his office or home, where the microgateway is located. The user may wish to place an international call, but doing so with his wireless connection may be prohibitive. In such a case, the user may call his home or office microgateway and then bridge a VOIP telephone call via the microgateway to the international destination. This can help reduce the overall toll for the call.

Three-Way VOIP Calling Using Peer-to-Peer Architecture

Also disclosed herein is a method for three-way calling between VOIP clients using no external “media concentration” or mixing device. Assume that two microgateways, a first microgateway MG1 and a second microgateway MG2 are in voice communication with each other, allowing a calling party to converse with a called party in a manner described above. This could be similar to the description of point-to-point communications between two VOIP-enabled CPEs.

When one of the two parties in conversation, say, the calling party, wishes to establish a three-way conference call, which may include a third party, the party sends a hook-flash signal to the microgateway to which it is connected. The microgateway then acts as a switch, and suspends the call temporarily and provides a dial tone to the party that sent the hook-flash signal.

At or about the same time, MG1 allocates a register or other service circuit and allocates a conference bridge circuit. Alternatively, the microgateway may initiate a second VOIP call from MG1 over the LAN or the dialup VOIP connection. Conference bridge circuits are similar to two-way voice channels, except that in these circuits, more than two channels converge at a single point, which could be the conference bridge service circuit. Data read from any one channel is copied to all other channels connected to the conference bridge.

A voice channel is created to establish a third leg of connection with the third party's equipment with the conference bridge circuit. A connection is set up with the third party's equipment, which includes transmitting an alert signal to the third party equipment (e.g., a ring tone) and receiving an answer signal from that third party equipment. These details are similar to tasks that a central office switch performs when establishing a three-way conference call.

After a third party is connected to the conference bridge circuit, data from each party is copied to the other two parties, thereby establishing a three-way communication. The same procedure may be expanded to include a multiple party conference circuit.

Enhanced VOIP Architecture for a V.34 Modem

Internet Service Providers, such as America Online or Fry's Internet, enable a customer operating a personal computer or other similar device to connect with the Internet either directly or via the ISP's network. Typically, these connections use a V.34 class of modems. Where data connections are established—for example, where facsimile and other connections are required—a data compression/decompression scheme known as V.42 is typically utilized. These features—V.34 for connection to the Internet, and V.42 compression/decompression for data transfer—typically are not optimal for delivering high quality VOIP telephone calls to a user's CPE, such as a telephone handset. Accordingly, there is a need to advance the art by enhancing a user's CPE.

Noting that the VOIP architecture uses a TCP/IP connection and a Point-to-Point protocol connection between the personal computer and ISP modems, the state of the art can be enhanced using the following method. First, error correction is turned off. Because VOIP is deterministic, based on a sampling or frame rate of 30 ms—if a packet is lost, this can be ignored instead of waiting for it or adding any delay to correct the error. Second, data compression, such as MNP, can be turned on to reduce the size of text sent over a connection giving an apparent increase in transfer speed. This is likely not the best solution for a VOIP call because voice signal is already coded in an optimized way. Additionally, data compression designed for data transfer must be turned off. Third, the microgateway is connected at a specific predetermined baud rate. It has been found that 21600 bits per second is suitable for a low symbol rate without impacting the quality of service. Because a VOIP call may need about 17 kb/s to connect, 21,600 b/s is appropriate. These steps reduce the code size, MIPS, and complexity while increasing the performance of the modem for VOIP applications. This method could be used potentially on a variety of modems, such as V.32 14.4 kbps modems and V.90, as well as with ADSL systems especially when these devices operate in voice-mode.

MG1 registers with server once turned on. After that a periodic “heartbeat” packet is sent to the server to indicate that the device is still on line and alive, ready for a call. If a call is placed to another MG1, the server gets a packet from MG1 requesting the call to the other MG1. The server will look up a destination unit, determine if the destination unit is online, then signal the called and calling units. Once done, the calling and called units connect directly over the Internet for peer-peer communications. The server is signaled once the call is terminated and both MG1's are ready for a call. If a call is placed to a PSTN phone, MG1 connects to a H.323 gateway to setup the call.

VOIP Menu Display on Existing LCD Screens

A number of proprietary VOIP systems need to provide an acceptable and uncomplicated user interface to a user that wishes to program the VOIP system. For example, Aplio's proprietary system requires a user to use the DTMF keypad of a POTS telephone to enter data in order to program the system. Typically, a user is prompted to enter a dial-up telephone number for an ISP, a user name and a password. These data items are then repeated to the user via the earpiece of the telephone handset or through a speaker embedded within the system.

Another architecture could be to provide a keypad and an LCD screen within the VOIP system itself. But this increases the overall system cost. To overcome this issue, one may program the VOIP system to use—instead of or in addition to the keypad of a POTS telephone handset—an LCD screen that typically comes with a telephone handset or with a Caller-ID box. Such boxes are marketed by most of the telephone consumer electronic equipment manufacturers such as CIDCO, Inc. of San Jose, Calif.

A telephony gateway, CPE Control protocol and a telecommunication protocol engine and method have now been illustrated. As described herein, the telecommunication protocol engine and method results in significant reduction in on-chip memory requirements and permits the use of otherwise memory limited microprocessors. It will also be understood that aspects may be embodied in other specific forms without departing from the scope of the disclosure and that the examples and embodiments described herein are in all respects illustrative and not restrictive. Those skilled in the art of the present disclosure will recognize that other embodiments using the concepts described herein are also possible.

Claims

1. A method comprising: receiving a dialing code from a calling device at a customer premise device coupled to the calling device;determining a network of a called device based on the dialing code;selecting a telecommunication protocol based at least in part on the network of the called device; andinitiating a call to the called device using the selected telecommunication protocol.

2. The method of claim 1, wherein, when the dialing code comprises a predetermined prefix, the network of the called device is a same network as a network of the calling device and the call is initiated as an on-network call.

3. The method of claim 1, wherein the dialing code is preceded with and followed by a first non-numeric symbol.

4. The method of claim 3, further comprising receiving a personal identification code from the calling device.

5. The method of claim 4, wherein the personal identification code is separated from the dialing code by a second non-numeric symbol that is different from the first non-numeric symbol.

6. The method of claim 1, wherein initiating the call comprises interacting with a service provider network device, wherein the service provider network device performs call set up signaling on behalf of the calling device to establish the call, and wherein the calling device and the called device communicate without the service provider network device after the call is established.

7. The method of claim 1, wherein initiating the call using the selected telecommunication protocol comprises: selecting a telecommunication protocol template by a processor of the customer premise device, wherein the telecommunication protocol template includes one or more virtual machine instructions executable by a virtual machine at the customer premise device to implement the selected telecommunication protocol;reading a first virtual machine instruction of the telecommunication protocol template from a first memory device of the customer premise device;initializing a first finite state machine using the first virtual machine instruction, wherein the telecommunication protocol template includes first template state data defining a first template state of the first finite state machine; andstoring the first template state at a second memory device of the customer premise device in a first state table, wherein the second memory device is internal to the processor.

8. The method of claim 7, further comprising: receiving an input at the processor via the call;determining, by the virtual machine, an updated template state of the first finite state machine based on the first state table and the input; andstoring an updated first state table at the second memory device, the updated first state table specifying the updated template state.

9. The method of claim 7, further comprising: receiving a request at the processor to implement a second telecommunication protocol;reading a second virtual machine instruction from the first memory device;initializing a second finite state machine using the second virtual machine instruction; andstoring a second template state of the second finite state machine at the second memory device in a second state table.

10. The method of claim 7, wherein the first memory device is a FLASH memory device and the second memory device is a RAM memory device.

11. The method of claim 1, wherein the selected telecommunication protocol includes at least one of a Session Initiation Protocol (SIP), a H.323 protocol, a STUN protocol, and a dynamic host configuration protocol (DHCP).

12. A method comprising: receiving a signal from a telephony device coupled to a customer premise equipment telephony gateway during a call between the telephony device and a remote telephony device via the customer premise equipment telephony gateway;in response to the signal, interrupting the call at the customer premise equipment telephony gateway;initiating a second call to a second remote telephony device from the customer premise equipment telephony gateway;sending information from the customer premise equipment telephony gateway to a second telephony gateway, wherein the second telephony gateway is a customer premise device coupled to the remote telephony device, wherein the information notifies the second telephony gateway of the second call; andsharing data among the telephony device, the remote telephony device and the second remote telephony device via the call and the second call.

13. The method of claim 12, wherein the signal comprises a hook flash.

14. The method of claim 13, wherein the hook flash enables access to a control protocol of the customer premise equipment telephony gateway via the telephony device.

15. The method of claim 12, wherein initiating the second call comprises: selecting a telecommunication protocol template by a processor of the customer premise equipment telephony gateway, wherein the telecommunication protocol template includes one or more virtual machine instructions executable by a virtual machine at the customer premise equipment telephony gateway to implement a telecommunication protocol;reading a first virtual machine instruction of the telecommunication protocol template from a first memory device of the customer premise equipment telephony gateway;initializing a first finite state machine using the first virtual machine instruction, wherein the telecommunication protocol template includes first template state data defining a first template state of the first finite state machine; andstoring the first template state at a second memory device of the customer premise equipment telephony gateway in a first state table, wherein the second memory device is internal to the processor.

16. A communication system comprising: a service provider gateway configured to communicate with one or more telephony gateways, including a first telephony gateway and a second telephony gateway, via a switched packet network, the service provider gateway comprising: a network communication system to communicate via the switched packet network;a protocol handling system to process data communicated according to a selected telephony protocol;a bridge system to connect the first telephony gateway to the second telephony gateway in response to data received from the first telephony gateway; anda billing system to facilitate billing users.

17. The communication system of claim 16, wherein the service provider gateway is further configured to communicate with one or more telephones via a public switched telephone network.

18. The communication system of claim 16, wherein the service provider gateway further includes an interactive voice response system to automatically respond to billing questions.

19. The communication system of claim 16, wherein the service provider gateway further comprises an authentication system to authenticate a user of the first telephony gateway.

20. The communication system of claim 16, wherein the service provider gateway is configured to determine whether a call initiated via the first telephony gateway is an on-network call based on a portion of a dialing code received at the first telephony gateway.