WO2000024166A9

WO2000024166A9 - Dynamic ip address assignment

Info

Publication number: WO2000024166A9
Application number: PCT/US1999/024651
Authority: WO
Inventors: Henry Hunyih Houh
Original assignee: 3Com Corp
Priority date: 1998-10-16
Filing date: 1999-10-18
Publication date: 2000-09-14
Also published as: AU2254900A; WO2000024166A1

Abstract

In a data network system (31, 70, 36, 30, 10') and/or process for communications between endstations (15', 17) which are routed through an IP network (30, 31) precious and distinctive addresses, typically IP addresses are not permanently associated with each endstation, but rather are only transiently related to some relatively versatile endstations (15, 17). Precious IP addresses are collected, dispensed and assigned to versatile endstations from a subnetwork-based pool of IP addresses (24, 28) according to network call like cycle events. At the start and then at the end of any call between a versatile end station and any other end station, a network controller (20) will perform assignment/dispensation from, and then return to, a pool of unique IP addresses such that even a versatile endstation has a unique IP address for the duration of the call.

Description

D YNΛMIC IP ADDRESS ASSIGNMENT

CROSS REFERENCE TO RELATED APPLICATION

This is a non provisional application claiming the benefit of provisional application 09/325,906 filed October 18, 1998.

FIELD OF THE INVENTION

The invention relates in general to IP address assignment as well as a telephone

system which is operated using a data network, such as a local area network and more

particularly to the use of such a system in both level 2 and level 3 network environments.

BACKGROUND OF THE INVENTION

Computer networks or data networks connect a plurality of devices to each other using a network architecture. Most network architectures provide several different layers.

Each layer is responsible for providing some service to the layer above and does this by

using the services of the layer below. The open systems interconnection (OSI) reference

model defines seven layers for computer networks. There is no special importance as to

the number seven. However the reference model provides some guidance for designers.

Protocols that are used for the various levels have evolved over time and some of the various layers have been subdivided into further layers.

Generally the first or lowermost level is referred to as the physical layer. This layer

has responsibility to transmit unstructured bits of information across a link of the network. The physical layer deals with such problems as size and shape of connectors, assignment

of functions to pins, conversion of bits to electrical signals, and bit-level synchronization.

It is usual for several different types of physical layers to exist within a network and even

for multiple different types of physical layers to exist within a node (also referred to herein

as an end station or end point device), as each technology requires its own physical layer.

For example, a node with an Ethernet attachment and an attachment to a 56Kb

synchronous line will have implemented two different physical layers.

Layer two is generally considered the data link layer which has the responsibility

to transmit chunks of information across a link. Level two deals with such problems as check summing to correct data corruption; orderly coordination of the use of shared media, as in a LAN (Local Area Network); and addressing when multiple systems are

reachable, as in a LAN. The addressing is accomplished with so-called MAC (Media

Access Controller) addresses. Specifically, each networkable device has assigned to it a

unique MAC address for use at the so-called layer two. Devices can communicate with each other based on the MAC addresses. Data packets may be switched based on MAC addresses. It is common for layer two links to implement different data link layers and for

a node (or end point) to implement several data link layer protocols, one to support each

of the different types of links to which the node is attached (as discussed above with

regard to layer one).

Layer three is normally referred to as the network layer. Layer three has the

responsibility lυ enable any pair of systems in the network to communicate with each other, A fully connected network is one in which every pair of nodes has a direct link

between them. However, this type of topology is not used as it does not scale beyond a

few nodes. Accordingly, in a more typical case, the network layer must find a path through

a series of connected nodes, the nodes along the path must forward packets in the

appropriate direction. The network layer deals with such problems as route calculation,

packet fragmentation and reassembly (different links in the network have different

maximum packet sizes), and congestion control.

With the more frequent use of the Internet, Internet protocol (IP) addressing has

been more extensively used at layer three. Routers and other layer three devices typically have address lookup tables wherein a packet which has an IP encapsulation (namely an IP address added to the packet) can be directed or routed by a router (or a network of

routers) based on the use of a lookup table of route entries which represent individual IP addresses and groups of IP addresses -often bit contiguous (there is a commonality

between leading bits of addresses).

Computer and telephone networks have historically been provided based on

separate physical infrastructures and are normally separately managed. Computers which are connected to the global Internet require IP addresses in order to communicate with

other computers around the world. For this reason, layer three devices often use layer three IP addressing. These same computers can communicate on local networks without the need for IP addresses by using layer two switching using MAC addresses. However, typically, IP addresses are used and layer three routing and interconnection is provided.

Telephone systems have typically been provided as PBX systems or similar systems

with line cards or other connections to the public phone system and with various

telephones connected back to a central exchange device. The PBX systems include digital systems wherein a proprietary protocol or other some phone-based protocol is used. With

such systems, most telephones do not have an IP address. Trying to converge the

infrastructure such that the telephone system operates over a computer network poses some challenges, particularly with regard to addressing. If a company were to replace its telephone system with a new IP-based phone system, they would need to double (or more)

the number of IP addresses they use. Thus, efficient management of the limited IP version 4 address space is an important consideration for such a converged infrastructure.

An IP address allocation scheme referred to as DHCP (Dynamic Host

Configuration Protocol) is known. This protocol functions for environments which are

primarily at level three. With DHCP the devices lease an IP address for their primary method of communication. While such DHCP leases can be short term in nature, the lessee usually cannot do anything meaningful without the IP address. This presents the problem of not being able to have communication within the subnet based on MAC addresses or

the like.

Typically level three packets are encapsulated in IP (Internet Protocol) and may be routed by routers based on EP addresses. MAC addresses, which are globally unique,

may be used for switching at level two, namely switching based on MAC addresses.

Typically with IP encapsulation, the destination and source IP addresses are provided.

By utilizing level two functions, a system may be provided which is able to

communicate without an IP address and hence does not require an IP address for normal

usage. However as the traffic is directed to outside of the local net, it is necessary to use the router MAC address and then use IP encapsulation including both the IP destination address and IP source address. The DHCP system is not optimized for systems relying

primarily on level two addressing and a system which primarily uses level two addressing

presents the problem as to functioning in a routed level three environment.

SUMMARY AND OBJECTS OF THE INVENTION

It is an object of the invention to provide a system which operates using a computer network infrastructure wherein a plurality of devices at network endpoints are

connected via the network (a subnet) and at least some of the devices do not have Internet

protocol (IP) addresses. For traffic beyond the subnet, or other traffic that requires an IP

address, a network control processor (NCP) is provided which allocates IP addresses to

the device on an as-needed basis.

The devices connected via the network (a subnet) wherein at least some of the

devices do not have Internet protocol (IP) addresses are preferably telephone devices. The telephone devices may be e.g. phones, TLIMs (Telephone Line Interface Modules), PSTN (Public Switch Telephone Network) gateways, Tl gateways, H323 gateways, etc., computers etc. In this text, the term telephone device is intended to include a telephone unit that has a handset and has a transmitter and receiver function for placing audio level

two (2) packets (e.g., packets with MAC addresses) on a network and for receiving such

data packets. The telephone device may also be a computer that has a sound card for audio input and output and a connection to the network (e.g., via a network interface card

which includes the transmitter and receiver function for placing level two packets on the

network and for receiving such data packets). The telephone device may also be a TLIM

or other device for converting one format of telephone signaling to another form of

telephone signaling (e.g., proprietary) and may also be for audio format conversion (one packet format to another) and for converting an audio format to another -e.g., analog

acoustic to a digital packet).

The system of the invention eliminates the need for a large number of IP addresses to be allocated, namely an EP address allocated to each end point, such as each telephone device attached to a network. Instead, a pool of EP addresses is maintained wherein the

total number of EP addresses in the pool is preferably less than the total number of phone

devices connected to the subnet. The allocation on an as-needed basis reduces the

complexity and overhead of managing the IP address space.

According to the invention a system is provided, which has features for the

management of EP addresses. The system has a number of devices located on the same level two network (a subnetwork or cluster of devices which can communicate with level two addresses). At least some of the devices have no IP address. Occasionally, one or more of these devices needs to communicate with a device located on a different subnet (an IP subnetwork) or a device with an EP address on the subnet. Communication between

two of the devices using IP addressing may also be provided. One device of the subnetwork or "cluster" can be a controller for the subnet. The

controller may also be connected to the network. The device controller is either active in handing out IP addresses or only responds to IP address requests. The device controller

may or may not be the same device which controls other features and functions of the overall system (the system may include for example the cluster itself as well as other devices not located on the same level two network).

According to a preferred embodiment of the invention, the device controller or

network control processor (NCP) knows the status of all the devices in the system.

Specifically, the devices are connected to allow communication between devices. The NCP

can note the desire of any particular device to communicate with other devices in the system but which are not in the cluster, or in a wholly separate system (which may or may

not be on the same level two network but are connected to the cluster). The system

preferably provides that the devices direct requests to the NCP. Upon the receipt of such a request, the device controller allocates to the device an EP address from a pool of IP addresses designated for the cluster. Such a pool of IP addresses may be maintained in a memory associated with the device controller. Upon notification that the device no

longer needs to communicate outside the cluster, the device controller reclaims the IP

address and returns the address to its pool for allocation to another requesting device. The

device controller may use another protocol (DHCP) from another controller or DHCP server to allocate this pool of IP addresses. The provision of the EP address can be

automatic or manual DHCP can be used by the NCP to obtain multiple EP addresses

wherein the lease for such addresses can be renewed as needed.

The preferred embodiment of the invention is thus able to allow communication without an IP address, within the cluster and hence does not require an IP address for

normal usage. In this normal function, the system communicates exclusively via level two

protocols when possible, and only uses level three protocols when necessary. If a level

two device (with no IP address and only a level two address) needs to communicate with

a level three device, the level two device is assigned an IP address for the duration of the call. The EP address is revoked at call termination or at some point after the call terminates. Accordingly, the pool of EP addresses may be maintained small and the number of IP address space resources required for the overall system is a function of the maximum number of expected calls between routed networks (e.g., the number or requests for communication between a device on the subnet with no IP address and a device with an

EP address and connected to the subnet), and is not a unitary function of the number of

devices in the system. As a single device controller or network control processor can be

designed to be switched at level two, then EP addresses are needed only when devices are making inter-domain calls (calls to be routed using a level three router to a different domain).

According to another embodiment of the invention, the device itself (e.g., phone

or phone system in a computer) recognizes that it needs to communicate outside of the cluster, and makes a request to the device controller for an IP address. The device uses the IP address received for the duration of the communication, and releases the IP address

back to the device controller at the end of the call. The device may or may not use an

existing protocol such as DCHP to grab an EP address. Protocols such as DHCP can be

used and allow the device controller (DHCP server) to be located on a different subnet than the entire cluster itself. In that case (no DCHP server on the same subnet) software known as BOOTP relay agent can be installed or activated on a router to relay the level 2 DHCP request or level 2 request to the appropriate DCHP server or domain controller. According to another aspect of the invention, a process for allocating IP addresses

is employed with a system with devices which normally communicate at level two and

require an IP address for communication via an IP router to another cluster or

subnetwork. The system utilizes a plurality of phone devices without EP addresses and a

Network Control Processor (NCP) which controls level two communication between the devices. The process includes detecting when a phone (A), without an EP address, goes off hook. A level two packet is sent to the NCP, informing the NCP of the off hook state

of the phone (A). A number for another phone (C) is dialed at the phone (A). The digits dialed are sent as a level two packet to the NCP. When the NCP detects that the phone (A) without the IP address and the phone (C) corresponding to the number dialed are not on the same level two network (and the phone dialed has an EP address but the phone that

dialed does not have an IP address) the NCP accesses an IP address from an IP address

pool maintained by the NCP. The pool is for use with devices on the same level two

network as the NCP. The NCP then sends a level two packet to the phone (A) with one

of the IP addresses from the pool and instructs the phone (A) to use the EP address for the duration of the call (e.g. A.A.A.A). The NCP also instructs the phone (A) to talk to the other phone (C) based on the known IP address (e.g., C.C.C.C). The phone (A) then grabs the IP address (e.g., A.A.A.A) and broadcasts an ARP (Address Resolution Protocol) message to the Local Area Network so as to advise the other devices on the

local subnet. The phone (A) then begins to send audio packets, encapsulated as IP packets,

to the other phone. The source IP address of the IP packets is e.g., A.A.A.A and the destination IP address of the EP packets is e.g., C.C.C.C. Upon completion of the call, either phone (A) or the other phone (C) hangs up. The phone (A) sends its information to

the NCP via a level two packet if it is phone (A) or via a level three EP packet or level two

packet if it is the other phone (C). The NCP upon receiving the packet indicating the termination of the call instructs the phone (A) to terminate the call and stop sending IP

audio packets to the other phone. The NCP also instructs the phone (A) that it no longer

has the EP address which has been allocated.

The various features of novelty which characterize the invention are pointed out

with particularity in the claims annexed to and forming a part of this disclosure. For a

better understanding of the invention, its operating advantages and specific objects attained by its uses, reference is made to the accompanying drawings and descriptive matter in which preferred embodiments of the invention are illustrated.

BRIEF DESCRIPTION OF THE DRAWINGS

In the drawings: Figure 1 is a diagram showing system components according to the invention;

Figure 2 is a diagram showing the system according to the invention with a level

2 packet exchange between devices in a subnet or cluster;

Figure 3 is a diagram showing aspects of a call setup for an exchange of packets between telephones using level 2 addressing and level 2 protocols;

Figure 4 is a diagram showing a call setup using dynamic Internet protocol address

assignment according to the invention;

Figure 5 is a diagram showing a call setup with a temporarily assigned EP address; Figure 6A is a flow diagram for illustrating steps involving the assignment of an

EP address; and

Figure 6B is another flow diagram illustrating steps involving the assignment of an EP address.

5

DESCRIPTION OF THE PREFERRED EMBODIMENT

Referring to the drawings in particular, the invention comprises a network system and process involving a plurality of interconnected devices defining a data network. The network is referred to as a subnetwork or cluster 10 and includes a physical connection

10 12 (such a connection may also be based on wireless interconnection schemes such as

radio frequency RF connections and infrared (IR) connections) between end points 14.

The end points include devices, which may be for example telephone units, computers,

TLIMs or other telephone devices 15. The end points also may have other devices 17 besides telephone devices. The devices may include a network interface with a transmitter

15 and receiver. The phone devices 15 also include a processor and also generate audio data

packets. At least some of the devices 15 and 17 have no EP address. Preferably all of the devices have a MAC address for communication over the network 12 with level two

packets (e.g., destination and source address). The end points 14 may have devices for

interconnection to other subnetworks or networks such as routers 30, bridges or switches.

0 The cluster 10 according to a preferred embodiment of the invention also includes a

network control processor (NCP) 20. The Network Control Processor 20 monitors traffic

over the network 12 and/or receives packets from other devices. The network 12 can implement any one of numerous protocols such as ETHERNET (EEEE802.3 using for example 10 base T or other physical media schemes). Although the preferred embodiment

of the invention is based on an ETHERNET type network, the invention can be practiced using various physical layers and various layer 2 protocols. The preferred embodiment relates generally to a Local Area Network as the cluster 10 however other networks

including Wide Area Networks, networks established based on the public phone system

and/or the Internet they also employ the system. However, the invention primarily provides the cluster 10 wherein most communication between the end points 14 may take place using level 2 addressing (e.g. the MAC addresses) associated with each device or end station at each end point 14. Further, the invention is not limited to a phone system

using the subnet. The invention also applies to other devices with a MAC address and

with no EP address, which can send and receive level two packets (e.g., packets using

MAC addressing).

The NCP 20 monitors traffic over the network and controls communication

between end points 14 which involves audio communication, namely telephone

communication with data packets. Besides the telephone communication using level 2

packets, exchanged between end points 14, the network interconnection 12 may be used for data exchange between computers, using the ETHERNET protocol as mentioned

above.

The NCP 20 includes a processor or an intelligent device 22 as well as a transmitter and receiver 26 and a memory 24. The memory 24 establishes an IP (Internet Protocol) address pool wherein a plurality of IP addresses are maintained or accessed. The IP addresses are first obtained in a known manner and input into the memory as

shown at 28. The number of IP addresses which are provided in the memory IP address pool 24 depends upon the anticipated or expected calls between an end point 14 and an

end point or device which is not on the local subnet or cluster 10.

The devices at end points 14 include telephone units which have no IP address.

These telephone units do have a MAC address which facilitates layer 2 communication between any of the various phones at end points 14 and the NCP 20. As shown in Figure

2 a telephone device A with no IP address can initiate a telephone call to device B which

also has no EP address wherein these devices are connected over the network connection

12 of the local subnet or cluster 10. To set up the call the NCP 20 sends a packet on the network connection 12 which includes the device A MAC address and signals device A

to talk directly to device B using device B's level 2 address. Similarly, the NCP 20 signals device B to talk directly to device A using device A's level 2 address. The packet

exchange between device A and device B occurs at level 2 wherein the packets include the

MAC addresses, namely the destination and source MAC address.

As shown in Figure 4, the system of the invention also allows communication

between a phone 15 or device 17 at an end point 14 and a phone or device 36 connected

to another network, subnet or the like wherein the subnet or cluster 10 is connected to the

other network or subnet via a level 3 router (EP router) 30 or routers 30 or a network of

routers 31. Specifically, the EP router 30 as shown in Figure 1 is a level 3 device which

for example may maintain a lookup table of EP addresses or groups of EP addresses for

determining where a packet is to be forwarded to. Normally, the term router refers to a device which can handle level 3 addresses. Most typically, the level 3 addresses use the

so-called EP (Internet Protocol) addressing. Switches also provide a similar function and level 2 switches are known which provide switching using level 2 addressing. Traffic to an entity outside a devices subnet 10 is provided with the MAC address to the router 30

with an IP address inside the packet. The router then can encapsulate the packet with the

destination EP address and source EP address. The EP router may also be considered a level

2/level 3 interface. Devices know whether the destination of a packet is on the same subnet or different subnet based on a subnet mask which can be maintained by the intelligent devices. A function of a source device IP address, a source devices subnet mask, a destination devices EP address is to indicate whether or not a destination is on the same subnet, level two network or logical level 2 subnet as the source. When a device communicates to another device on a same subnet using IP an ARP request is generated.

The source device responds with its own hardware address (MAC) and the two devices

can communicate at level 2 or level 3 as both source and destination have MAC and IP

address. The IP router has knowledge that some end-point or device on the subnet has an EP address corresponding to the received ARP broadcast. The router or destination

device fills in its own hardware address and responds to the requesting device. It may also put the hardware address in its own ARP table. The ARP request involves a response of

devices on the same net. Also, intermediate devices may make proxy responses for devices not on the net. Virtual LAN concepts can be used with the system of the

invention.

As shown in Figure 4, where a device with no IP address such as device A at end

point 14 wishes to set up a call with a device on a different subnet or connected via the

level 3 interface (the IP router 30) the NCP 20 must first assign it one of the IP addresses from the IP address pool 24. The call setup is shown in the diagram of Figure 5. The NCP 20 assigns device A with a level 3 address by sending a level 2 packet to device A. NCP 20 then signals device A to talk to device C using device C's level 3 address.

Similarly, the control unit signals to device C to talk directly to device A using device A's

temporary level 3 address. The packet exchange between devices A and C occurs at level

3 via the router 30 (or network 32 with a network of routers, subnets etc.). When the call

is done, the NCP tells device A and C to terminate the call. Then, the NCP 20 revokes

device A's level 3 address.

As to the router 30, the router may assign the temporary IP address to a particular

devices MAC address in its ARP lookup table, but this may be changed during subsequent

calls. Figure 6A shows a flow diagram of process steps involved in a call which requires the assignment of an EP address as discussed above.

The process of the invention is initiated at 60 as the phone A, namely a device 15

in the subnet or cluster 10 which has no EP address, has its status changed to off-hook.

This may be for example by lifting a handset or otherwise actuating the phone A. Phone

A is at an end point 14 connected via network connection 12 and provided in a subnet or

cluster 10. The change of status to off-hook results in level 2 packets being sent to the

NCP 20 informing the NCP 20 that phone A is off-hook. This is shown in the flow

diagram at number 62. Number 64 shows the subsequent state wherein phone C (for example with number 234) is dialed on phone A. This results in the digits being sent in level 2 packets to the NCP 20. The subsequent step 66 is shown wherein the NCP 20

knows that phone A (at number 123) and phone C (at number 234) are not on the same

level 2 network. The NCP 20 knows that phone C (at number 234) already has an EP

address but that phone A (at number 123) does not have an EP address. The subsequent step at 68 involves the NCP accessing an EP address from the address pool 24. The

processor 22 can use any one of a number of algorithms for accessing the IP address including accessing the next available EP address. Another algorithm can be implemented

if there are no IP addresses available. However, typically a number of EP addresses are available and the NCP signals to the EP address pool 24 to read out an EP address from

memory which is to be assigned to one of the devices on the same level two network as

the NCP 20. At the subsequent step 70 the NCP 20 sends a level two packet to phone A with an EP address read out from the IP address pool (e.g. A.A.A.A) and instructs

phone A to use this IP address for the duration of the call. At the subsequent 72 step the

NCP 20 instructs phone A to talk to phone C (number 234) which is at IP address C.C.C.C. (see also Figures 4 and 5). At the subsequent step 74 phone A (number 123)

grabs the EP address A.A.A.A and advises the local subnet (cluster 10) by broadcasting an

ARP message to the local network. That is, a level 2 packet is sent addressed to each end point 14 of the subnet 10 using the address resolution protocol (ARP). At the subsequent

step 76 the phone A sends audio packets encapsulated as internet protocol packets to

phone C (at number 234). The source LP address of the IP packets is A.A.A.A and the

destination IP address of the EP packets is C.C.C.C. This is received at the interface or IP router 30 which forwards the packets to the subnet 50 based on the destination IP address. In the opposite direction the phone C sends audio packets encapsulated as EP packets to the phone A (at number 123). The source EP address of the packets is C.C.C.C

and the destination EP address of the EP packets is A.A.A.A. Based on the address

resolution protocol broadcast the router 30 knows that an entity on the subnet or cluster 10 has the IP address of the earlier ARP broadcast. Packets are exchanged during the phone conversation as shown for example in Figure 5. Subsequently the process continues

to step 78 wherein either phone A or phone C hangs up. The phone that hangs up sends this info to the NCP 20 via a level 2 packet or via a level 3 EP packet (in the case of phone

A) or a level 3 EP packet (in the case of phone C). The NCP 20 then instructs phone A

to terminate the call, to stop sending EP audio packets to phone C and it indicates that it no longer has the EP address A.A.A.A. This last step is shown at 80 in Figure 6.

The system of the invention also allows a phone device with an IP address to call a phone device or other device (15,17 etc.) which has no EP address. The process is similar to the process described with reference to Figure 6A. As shown in Figure 6B a

phone C (x234) with an IP address goes off hook at shown at 82. As indicated at 84,

phone C (x234) sends level three packet to NCP 20 informing the NCP that phone C is off hook. Phone A is dialed at 86. The digits are sent in level three packets to be NCP

20. The NCP 20 knows that phone C has an IP address and knows that phone A (xl23) has no EP address as indicated at step 88. Next, the NCP 20 grabs an IP address from the

address pool 24 as indicated at 90. As shown at 92 the NCP 20 sends a level two packet

to phone A with the EP address (e.g. A.A.A.A) and instructs phone A to use this IP address for the duration of the call. At 94 the NCP sends a level 2 packet to phone A instructing phone A to talk to phone C which is at a particular IP address (e.g., C.C.C.C).

Phone A grabs the IP address and advises the local subnet by broadcasting ARP messages

to the local network as indicated at 96. Phone A and phone C exchanged audio packets

encapsulated as EP packets as indicated at 98. Either phone A or phone C terminates the call as indicated at 100. The NCP 20 instructs phone A to terminate the call, to stop sending IP audio packets to phone C and that it no longer has the address that was assigned (e.g. A.A.A.A) as indicated at 102.

The process of the invention for using the system of the invention can also provide

phone devices on the same subnet with IP addresses for communication using level 3

packets. The invention is not limited in any way to the phones or other devices being on

different subnets. Either or neither of the phone devices or other devices may have no IP address. The NCP may assign in EP address to either phone device or other device. Even

though it is advantageous to provide communication with level two packets, communication on the same subnet may be provided with level three packets.

According to another embodiment of the invention the system includes devices which can be in a single cluster 10 or can be distributed (in multiple clusters or individual devices or a combination thereof), all logically associated with the same NCP 20 (see

Figure 1). For example, the level 3 device 37 in Figure 1 can be logically part of a system

70 controlled by NCP 20.

According to yet another embodiment of the invention, the system of the invention

can be used in two wholly separate systems controlled by two separate NCPs 20, 20',

respectively, wherein a phone device 15 in one system which does not have an IP address may wish to contact a phone device 15' in a different system or subnet 10' which also does

not have an EP address. The NCP 20 follows a procedure in which:

1) The calling phone device 15 is activated and dials a number which indicates the destination phone device 15' directly (unified/universal dial plan), or which maps to a specific system (system code) and then to the phone device 15' on that system

(system- specific extension).

2) The NCP 20 recognizes that the number dialed is not a phone device within its system, and determines the system which controls the destination phone device 15' (either

by looking up the system code in a local database or by contacting some external device which can perform the mapping).

3) In an exchange of signaling messages, the NCP 20 for the system for the calling phone device contacts the NCP 20' for the system of the destination phone device and

indicates to the other NCP 20' that the source device 15 is trying to reach the destination

phone device 15'.

4) Prior to or during this exchange of signaling messages, the NCP for the system of the source (calling) phone 15 assigns an EP address to the source phone 15, and passes this information to the NCP 20' for the system of the destination (called) phone device 15'.

5) During the exchange of the signaling messages, if the destination phone device

is activated, the NCP for the system of the destination device assigns the destination phone

device an IP address and passes this information back to the NCP for the system of the calling device.

6) If the destination phone device 15' does not answer and the source phone

device 15 also hangs up prematurely, an IP address may or may not be assigned to the

source phone device 15. The preferred method is that the IP address assignment happens

when the call is connected to the destination phone device 15' or audio recording device

(i.e., voicemail on the destination system). While specific embodiments of the invention have been shown and described in detail to illustrate the application of the principles of the invention, it will be understood that the invention may be embodied otherwise without departing from such principles.

Claims

WHAT IS CLAIMED IS:

1. A system for data networks, the system comprising:

a network interconnecting a plurality of end station devices, each end station device having a level 2 address and some or all of said end station devices having no assigned Internet protocol address;

a network controller connected to said network interconnection, said network controller, said network interconnection and said end station devices cooperating to define

a subnetwork with devices communicating in the subnetwork using level 2 packets; an IP address pool associated with said network controller, said EP address pool

corresponding to a plurality of IP addresses, said network controller assigning one of said

IP addresses to one of said phones upon a commumcation being initiated which requires

or involves level 3 addressing.

2. The device according to claim 1, wherein said all of said end station devices

having no assigned Internet protocol address are telephone devices.

3. The device according to claim 1, wherein the number of EP addresses provided by said EP address pool is less than a number of said some or all of said end station devices

having no assigned Internet protocol address.

4. The device according to claim 1, wherein said network is a subnetwork and

wherein said assigning of an EP address by said controller occurs upon a communication being initiated which involves a device on said subnetwork and a connection by an

interface to a device not on said subnetwork.

5. The device according to claim 4, further comprising a level 3 interface connected

to said subnetwork for providing level 3 connections between said subnetwork and a network or network devices connected to said interface.

6. The device according to claim 1, wherein the network controller establishes telephone calls between phones as end station devices based on packets exchanged

between end station devices with level 2 MAC addresses only.

7. The system according to claim 1 , wherein the network interconnection provides

an ETHERNET type network connection.

8. A system according to claim 1, wherein at the termination or subsequent to the termination of the call the device which had been assigned the EP address from the EP

address pool no longer has the IP address.

9. A process for operating a telephone system using a data network with a

network connection between end points with phone devices connected at respective end points and a network controller connected to the network, the phone devices and network

controller each having a unique MAC address, the process comprising: providing telephone communication between phone devices using the controller to initiate the sending of packets between devices which are to communicate, using MAC

addresses only;

providing an EP address pool which can be addressed by said network controller, said EP address pool having a number of EP addresses which is less than a number of phone devices connected to the network; and

upon a phone device connected to the network initiating a phone call to a device

not connected to the network, assigning one of said plurality of IP addresses to the phone device.

10. The process according to claim 9, wherein said step of initiating a phone call to a device not on the network include sending a packet with a MAC address only from

a phone device which goes off hook, informing the network controller of the off hook

status;

dialing at the network phone device a number which corresponds to a device which is not on the subnetwork resulting in sending a level 2 packet to the network controller

indicating the number of a device not on the subnetwork;

and at the network controller determining that the phone device which has dialed and the phone device dialed are not on the same level 2 network and determining that the phone device dialed has an IP address and the phone device which has dialed does not have an EP address.

11. The process according to claim 9, wherein a phone device with an EP address

calls a phone device without an IP address and the process includes generating a broadcast ARP (Address Resolution Protocol) message.

12. The process according to claim 9, wherein said step of accessing the IP address pool includes obtaining an EP address from the address pool and sending a packet with a

MAC address only from the network controller to the phone device, which packet includes

the IP address for instructing the phone device to use the IP address for the duration of the

call and instructing the phone device to communicate with the other phone device which has a particular IP address.

13. The process according to claim 9, wherein a phone device with an EP address calls a phone device without an EP address and the process includes generating a broadcast ARP (Address Resolution Protocol) message.

14. The process according to claim 9, further comprising upon receiving an IP

address packet from the network controller, the phone device broadcasts an ARP message

to the local subnetwork indicating that the phone device has the IP address which the

network controller has assigned to it; and subsequently sending audio packets,

encapsulated as IP packets to the other phone device wherein the IP source address is the address assigned to the phone device and the destination EP address is the IP address of the other phone device.

15. The process according to claim 9, wherein each phone device has no IP address

such that two EP addresses are assigned by the NCP for a call and the process includes generating a broadcast ARP (Address Resolution Protocol) messages for each phone

device.

16. The process according to claim 9, wherein upon one of the phone device and the other phone device hanging up, the phone device which hangs up sends a packet to the network controller wherein the packet is a level 2 packet of sent by the phone device and

the packet is a level 3 packet if sent by the other phone device.

17. The process according to claim 9, wherein upon termination of a call, the phone

device stops sending IP audio packets to the other phone device and the phone device no

longer has the EP address assigned to it by the network controller.

18. A process for operating a system using a data network with a network connection between end points with devices connected at respective end points and a network controller connected to the network, the devices and network controller each

having a unique MAC address and at least some of the devices not having a permanently

assigned IP address, the process comprising: providing communication between devices using the controller to initiate the sending of packets between devices which are to communicate, using MAC addresses

only;

providing an IP address pool which can be addressed by said network controller,

said IP address pool having a number of EP addresses; and upon a device connected to the network initiating a communication to a device wherein one or both devices have or require an IP address for communication or if

communication is desired using IP addressing, assigning an IP addresses from said pool

to one or both of said devices.

19. The process according to claim 18, wherein said step of initiating a communication to a device includes: sending a packet with a MAC address only, from a device for informing the network controller of the desire to initiate communication with another device; and

at the network controller determining that one or both of the devices to be involved in the communication has or requires an IP address or that the device dialed has an EP

address.

AMENDED CLAIMS

[received by the International Bureau on 18 February 2000 (18.02.00) ; original claim 16 amended ; remaining claims unchanged ( 1 page) ]

generating a broadcast ARP (Address Resolution Protocol) messages for each phone device.

16. The process according to claim 9, wherein upon one of the phone device and the other phone device hanging up, the phone device which hangs up sends a packet to the network controller wherein the packet is a level 2 packet if sent by the phone device and the packet is a level 3 packet if sent by the other phone device.

longer has the IP address assigned to it by the network controller.

18. A process for operating a system using a data network with a network

connection between end points with devices connected at respective end points and a network controller connected to the network, the devices and network controller each

assigned IP address, the process comprising:

providing communication between devices using the controller to initiate the

sending of packets between devices which are to communicate, using MAC addresses only;

providing an IP address pool which can be addressed by said network controller,

said EP address pool having a number of EP addresses; and

upon a device connected to the network initiating a communication to a device