3com Corporation

Utstarcom, Incorporated

Nurettin B Beser

John G Fijolek

Michael S Borella

Philip T Robinson

Ikhlaq S Sidhu

Guido M Schuster

David A Grabelsky

Ali Akgun

Levent Gun

Steven W Sangston

A system and method for Internet telephony between a caller station and a callee station are described. The caller station is connected to a first edge network via a first telephony interface, and the callee station is connected to a second edge network via a second telephony interface. An intermediate network is connected to the first edge network via a first router and is connected to the second edge network via a second router. The callee station is associated with a callee station number. The first router initiates the call in response to a setup message that includes the callee station number. A first gatekeeper, controlling the first router, and a second gatekeeper, controlling the second router, together mediate the process of setting up the call. A back end server, in communication with the first and second gatekeepers, stores the addresses and station numbers needed to set up the call. During the call, the first router performs network address translation to transmit signals between the first edge network and the Internet, and the second router performs network address translation to transmit signals between the second edge network and the Internet.

Internet telephony using network address translation

A method and application programming interface for using multiple network addresses on a common physical layer. The host protocol stack supports multiple Internet Protocol interfaces. When a process makes a function call to create a new socket, a new IP address is associated with the socket. Each socket is then bound to an IP address that is distinct from the IP addresses bound to other sockets. This is in contrast to conventional sockets that are bound to a common IP address. In this manner, each process may be associated with a unique IP address. Such a configuration may useful in Internet telephony where each call process receives a unique private IP address in a virtual private network.

Method and application programming interface for assigning multiple network addresses

A method to use the data packet carrying ability of cable TV networks as described in DOCSIS to accurately and quickly transmit a voice call from a cable user to another user over a cable TV cable. The cable telephone, or cable modem, is given the ability to detect when voice activity from the subscriber is above and below a predetermined value. When the cable modem has voice activity, the cable modem knows that it will have a continuous stream of voice data packets which need to be sent very quickly and very accurately to the CMTS. Therefore, the cable modem requests a periodic stream of time slots from the CMTS. When the cable modem detects no voice activity or a silence period from the subscriber, the cable modem indicates that the periodic stream of time slots is no longer needed and the CMTS stops providing the periodic stream of time slots. When voice activity of the subscriber resumes, the cable modem again request the periodic stream of time slots, and transmits the voice data packets or cells in these time slots.

Method to transmit silence compressed voice over IP efficiently in DOCSIS cable networks

A modem, communications system and method are provided with a transmitter and a receiver connected to a modem termination system (CMTS) by a medium. The modem has a memory forming a transmit queue including voice and video. The modem has a controller for sending data units from the receive queue to CMTS. The modem requests unsolicited grant service (UGS) from CMTS when the frame controller detects periodic stream to transmit. Upon detecting a data accumulation in the receive queue, the controller asserts a queue indicator. The Modem receives an occasional extra grant per interval and uses these grants until the queue indicator is not asserted. According to another feature, during transmission, the modem requests subsequent grants in a subsequent interval using an extended header having multiple bits for requesting the multiple grants. The CMTS allocates at least an extra grant to form a grant burst in response to the request.

Cable modem with dribble grant access system and method

A method and system for allocating persistent private network addresses between private networks. An Internet Protocol (&ldquo;IP&rdquo;) address of a multimedia device (e.g., a Voice-over-IP (&ldquo;VoIP&rdquo;) device) is publicly available and the device, the device's location or the device's user may be identified and become the target of a hacker. Persistent private IP addresses may be used for such multimedia devices. Persistent private IP addresses are unique and persistent for a duration of a multimedia session between two private networks. The persistent private IP addresses are not routable on a public network like the Internet. The persistent private IP addresses allow a first network device on a first private network to establish a data session with a second network device on a second private network over a public network such as the Internet. The first network device and the second network device negotiate a pair of private IP addresses that are persistent, private and unique across both the first private network and the second private network address space. The persistent private IP addresses may be allocated by Distributed Network Address Translation (&ldquo;DNAT&rdquo;) or Realm Specific Internet Protocol (&ldquo;RSIP&rdquo;) servers. The method and system allow a first private network to determine an estimated number of persistent private IP addresses to send to a second private network such that the second private network can select at least one pair of persistent private IP addresses that is not in use on the second private network with high probability of success (e.g., about 99.9% or greater). The method and system may improve security of multimedia data sessions (e.g., VoIP) and allow two private networks using private IP addresses to communicate over a public network like the Internet.

Method and system for allocating persistent private network addresses between private networks

A method and system is provided to send and receive Dynamic Host Configuration Protocol (&ldquo;DHCP&rdquo;) messages in a data-over-cable system via a route with one or more routers. The routers may apply one or more different protocol filters to DHCP messages, such as filtering out DHCP messages with a &ldquo;Martian&rdquo; Internet Protocol (&ldquo;IP&rdquo;) address (e.g., 0.0.0.0) and other characteristics. The Martian IP address is commonly used as an initial IP address in a DHCP initialization sequence but is often filtered out by routers as an invalid IP address. DHCP messages are received by a first protocol agent from a first User Datagram Protocol (&ldquo;UDP&rdquo;) port. The first UDP port is used by DHCP servers and network devices such as cable modems and customer premise equipment to send and receive DHCP messages via a route that may apply one or more DCHP filters. The first protocol agent sends the DHCP messages to a second protocol agent on a second UDP port via a route that does not apply filters to the DHCP messages. The first protocol agent receives DHCP messages from the second protocol agent on the second UDP port. The first protocol agent sends messages back to a DHCP protocol server or network device such as a cable modem on the first UDP port. The protocol agent allows DHCP messaging to be seamlessly used in a data-over-cable system with routers that apply a variety of protocol filters to DHCP messages.

Method and system for seamless address allocation in a data-over-cable system

A method and system for network address maintenance in a data-over-cable system. The method includes determining whether a network device, such as a cable modem, responds to a request in a timely manner. When updating network address tables, such as Address Resolution Protocol tables, an Internet Control Message Protocol (&ldquo;ICMP&rdquo;) echo message is generated and sent to the Internet Protocol (&ldquo;IP&rdquo;) address of a network device. If an ICMP echo reply message is not detected within a pre-determined time, the network address is deleted from the network address table. The deletion of the network address ensures that the device is isolated from the network. The method and system provide for the rapid removal of stale network addresses, freeing valuable memory space, and increasing security in a data-over-cable system.

Method for network address table maintenance in a data-over-cable system using destination reachibility

A method for network address maintenance in a data-over-cable system. The method includes receiving a message containing a first network address, such as an Internet Protocol address, to be stored in a network address table, such as an Address Resolution Protocol table. If the first network address equates to a second network address stored in the network address table, one or more network addresses associated with the second network address are deleted from the network address table. A third network address is then stored in the network address table. The method provides for the removal of stale network addresses, freeing memory space and increasing security in the data-over-cable system.

System and method for network address maintenance using dynamic host configuration protocol messages in a data-over-cable system

A method and system for provisioning network addresses in a data-over-cable system. Provisioning of network addresses allows multiple &ldquo;always-on&rdquo; network devices with multiple associated devices to be used on a data-over-cable system with a limited public network address pool. The &ldquo;always-on&rdquo; network devices provide services, such as, Voice over Internet Protocol (&ldquo;VoIP&rdquo;), that typically require instant access to data-over-cable system. Network devices such as &ldquo;always-on&rdquo; cable modems may allocated private network addresses (e.g., Internet Protocol addresses) on the data-over-cable system. The private network addresses are not addressable outside the data-over-cable system. Other network devices associated with the cable modems, such as customer premise equipment, may be allocated public network addresses (e.g., Internet Protocol Addresses) on the data-over-cable system. The public network addresses are addressable outside the data-over-cable system. The network address provisioning is accomplished by selecting a private network address marker and a public network address marker for selected network devices and using an extended Address Resolution Protocol table to determine a device type. The private network address marker or public network address marker is added to a Dynamic Host Configuration Protocol message field by a cable modem termination system. A Dynamic Host Configuration Protocol server uses the private or public address marker to allocate a private network address or a public network address on the data-over-cable system. The provisioning of network addresses allows two or more virtual networks to be created on the data-over-cable system, and may reduce a number of public network addresses required for a data-over-cable system.

Method and system for provisioning network addresses in a data-over-cable system

A method and system for network address maintenance in a data-over-cable system. The method includes determining whether a network device, such as a cable modem, registers in a timely manner. When updating network address tables, such as Address Resolution Protocol tables, a time value is associated with an address, such as an Internet Protocol (&ldquo;IP&rdquo;) address, of a network device. If a time between two registration events exceeds this time value, the network address is deleted from the network address table. The deletion of the network address ensures that the device is isolated from the network. The method and system provide for the rapid removal of stale network addresses, freeing valuable memory space, and increasing security in a data-over-cable system.

Method for network address table maintenance in a data-over-cable system using a network device registration procedure

The present invention relates to a method for optimizing data flow in a data network. Data that is to be transferred across the network at a constant bitrate is handled differently by a Transmission Control Protocol (TCP) layer compared to other data. A TCP congestion control process is adapted to recognize whether data packets from an application are to be transferred across the data network at a constant bitrate. The constant bitrate channel may be recognized by a Quality of Service identifier or a TCP port number. If the data packets do belong to the constant bitrate channel, the TCP congestion control limits the congestion window for these data packets. The congestion window for data packets other than those in the constant bitrate channel is allowed to increase, as usual, until data packets are lost, at which point the congestion window is reduced to a maximum segment size. The congestion window for constant bitrate packets, however, is increased until it reaches a maximum value. The limiting value of the congestion window is the constant bitrate multiplied by the round trip delay time between sending a packet and receiving its acknowledgement. The modifications to the TCP process may provide for the transmission of constant bitrate data over the data network that does not experience a timeout.

Method for optimizing data transfer in a data network

A method and system for quality-of-service in a data-over-cable system is provided. A cable modem in a data-over-cable system makes a connection request to a cable modem termination system with a requested quality-of-service. The requested quality-of-service includes class-of-service, quality-of-service and other related parameters. The connection request is sent from a cable modem or a cable modem termination system to a Quality-of-Service (&ldquo;QoS&rdquo;) server. The OoS server determines if the cable modem termination system has enough bandwidth to make the connection to the cable modem with the requested quality-of-service. If the cable modem termination system has enough bandwidth, a quality-of-service identifier is to returned to the cable modem termination system. The cable modem termination system uses the quality-of-service identifier to make a connection to the cable modem with the requested QoS to the cable modem. The QoS server reduces the computational burden and complexity of software on the cable modem termination system. The QoS server is flexible and adaptable to new QoS parameters and provides a standard way to balance QoS requests among multiple cable modem termination systems.

Method and system for providing quality-of-service in a data-over-cable system

A method and system for virtual network administration with a data-over-cable system. A virtual networking tag and a network address (e.g., an Internet Protocol address) assigned by a network other than a data-over-cable system is used to provide a virtual network to one or more network devices, such as cable modems via a data-over-cable system. The virtual networking tag may also be used to request a desired service class (e.g., Class-of-Service, Quality-of-Service, Type-of-Service, Service Level Agreements, etc.), for a desired end-to-end networking service (e.g., Voice over Internet Protocol). The virtual networking tag can be mapped to a data-over-cable service class. The data-over-cable service class can be mapped into a transport service class for a transport network. The virtual networking tag is used to provide a desired end-to-end networking service from a network device on a first external network (e.g., an Ethernet Local Area Network), through the data-over-cable system, through a transport network (e.g., Asymmetric Digital Subscriber Line, Asynchronous Transfer Mode, Frame Relay, Integrated Services Digital Network, Synchronous Optical Network, Voice over Internet Protocol, etc.) and to a second external network (e.g., another Ethernet Local Area Network). The method and system are used to provide a variety of networking services via and through a data-over-cable system.

Method and system for virtual network administration with a data-over cable system

A method and system for changing type-of-service with an associated quality-of-service in a data-over-cable system is provided. The method and system allow a cable modem or a cable modem termination system to dynamically override the statically assigned type-of-service supplied to the cable modem in a Dynamic Host Configuration Protocol (&ldquo;DHCP&rdquo;) initialization sequence. A selection input can request a second type-of-service that may or may not be supported by the cable modem on the cable modem termination system. If the requested second type-of-service with an associated second quality-of-service is permitted on the cable modem, the cable modem termination system dynamically changes the first type-of-service to the requested second type-of-service. Allowing a cable modem and cable modem termination system to change the statically assigned type-of-service to a second type-of-service allows the user greater flexibility for configuration cable modems in a data-over-cable system.

Growing community of inventors

Evanston, IL, United States of America

Nurettin B Beser