Electrical computers and digital processing systems: multicomput – Computer network managing – Computer network monitoring
Reexamination Certificate
1999-04-28
2003-01-14
Follansbee, John A. (Department: 2156)
Electrical computers and digital processing systems: multicomput
Computer network managing
Computer network monitoring
C709S220000, C709S227000
Reexamination Certificate
active
06507869
ABSTRACT:
TECHNICAL FIELD
This invention relates to a method and apparatus for tracking assets in a communications network.
BACKGROUND AND PRIOR ART
Asset Tracking Systems (ATS) are typically central components of inventory control and system management. Within a networked computer environment, ATS utilize client/server technology build on top of standard network protocols to query individual networked computers for a list of their assets and components which is then maintained in some inventory or asset database. However, the physical location of a computer system can not be queried dynamically through standard protocols and hence current ATS rely on people manually keeping the asset location data base current. Nevertheless, dynamically tracking the approximate physical locations of computers (particularly workstations) in an enterprise is a highly desirable feature. What is missing in today's tracking systems is the automatic identification of the physical location of the computer.
In a broader context the determination of the physical location of a device within its environment is primarily based on attaching sensors to devices and using radio frequency identification (RFI) as described in U.S. Pat. Nos. 5280159, 5785181, 5611051, 5608193, which are hereby incorporated by reference. Best known is the EAZY Pass System used at Toll Booths around the country. These methods are inadequate as they require that the sensor pass in close proximity of a detection device which initiates the radio frequency identification and receives the response.
In current local area networks (LAN) systems network attached devices, such as computer and printers, identify themselves to the other network attached devices through various standardized protocols. As an example of the various identification processes in today's network, we will describe an Internet Protocol (IP) based system. Here, a computer connected to a subnet finds the address of any other computer in the same subnet by following standardized address resolution and identification protocols.
Since this invention utilizes standard network protocols, it is imperative for the understanding of this invention to provide a review of the state of the art in networks and device identification as well as their limitation with respect to identifying the physical location of network attached devices. To that extent we incorporate the following literature on standard network protocols [1],[2] and [3] by reference.
Shown in
FIG. 1
is a typical LAN. A LAN is generally constructed of several subnetworks (
100
), each comprised of a set of devices, such as computers (
101
), printers (
102
), file servers (
103
) and splitters (
104
). The subnet (
100
) is typically connected to the rest of the network via gateway (
105
) and routers (
106
). Network protocols follow a layered approach, thus defining a protocol stack, shown in FIG.
2
. The bottom layer, or physical layer, is comprised of the physical network medium, and specifies the electrical characteristics, the connectors and the hardware carrying data. The data link layer utilizes this medium to define units of data, called frame or packet, consisting of frame header, frame data and optional frame trailer. Popular data link protocols are Ethernet, token ring, FDDI, ISDN, ATM, SLIP to mention a few. Layered on top of the data link are standard protocols such as IP, DecNET and NetBEUI, amongst others. The higher the protocol layer the more abstract are the services. For instance the lower layers often implement an unreliable communication medium, while higher level layers provide a reliable communication medium on top of unreliable mediums. The data-load of lower level packets typically embodies the entire packet of the next higher layer. For instance the Ethernet packet or frame, shown in
FIG. 3
, embodies as its data-load the IP packet, commonly referred to as the IP datagram. The IP packet format is shown in FIG.
4
. The IP datagram's data section embodies higher level packets defined for protocols such as UDP or TCPIP.
Referring to
FIG. 1
, within a network, network attached devices are referred to as hosts (
101
,
102
,
103
,
105
). Dependent on the layer of the protocol stack, hosts are addressed by different names or protocol addresses, short addresses. For instance, in a particular name domain a host might be identified by a more descriptive host name address such as “webserver.watson.ibm.com”. In the IP domain this host might be known by its IP address (e.g. 9.2.220.54). In the link layer the host might be identified by the 48-bit physical address of its Ethernet adapter. The physical address is often also referred to as the MAC (machine access control) number and is a globally unique number that is encoded into the network adapter at manufacturing time. In order to generate the packets as shown in
FIG. 2
, address translation has to take place in order to move from one layer of the protocol stack to a lower layer. For instance, domain name servers translate between host names and IP addresses. Note that a particular host can hold multiple addresses at a particular level. For instance, a host might have several communication adapters and therefore multiple physical addresses.
Let X and Y be two hosts attached to a subnet and further let IP(X) and IP(Y) be their respected IP address for a given physical protocol (e.g. Ethernet). If X wants to communicate with Y, then for the purpose of communicating over a particular physical layer, the IP address IP(Y) must be first translated into the physical address PA(Y) of Y for a particular physical protocol. For this purpose an address resolution protocol (ARP) module is consulted that caches translations between IP addresses and physical addresses. If such a translation exists, then IP(X) can send to IP(Y) by copying the PA(X) into the source and PA(Y) into the destination of the packet frame. If no such translation is available, the physical address PA(Y) must be discovered first. For this purpose the address resolution protocol (ARP) is utilized. Here, first a particular frame is broadcast on a subnet that contains the ARP request command token, the PA(X) as the source of the packet, a wild card target physical address PA(*), the higher layer source and destination addresses, which in the case of IP are the source IP address IP(X) and the target IP address IP(Y). Note, that the ARP is shared by all physical link protocols, such as IP, NetBEUI, etc. Since the ARP packet is broadcast with a wild card target PA(*), every host attached to the subnetwork, picks up the packet and pushes it up its communications stack. If the target IP(Y) address matches any computer on the subnet, this computer will respond with an ARP reply to the interrogator by means of exchanging the IP target and source addresses in the request packet, setting the physical source address to PA(Y) and the physical target address to PA(X). X will pick up this ARP reply packet, and store the translation <IP(Y),PA(Y)> contained therein in its ARP module. The translation is typically discarded after some time to ensure that translations are somewhat up to date. For instance, if a host disconnects from the network, packets addressed to this host will simply be discarded and higher level protocols will be notified of the communication error. A variant of the ARP called reverse ARP (RARP) helps a node find out its own higher layer address. A host can broadcast a query requesting its higher layer address and stating its physical address. A server that is configured with a physical to higher layer address table responds to such queries by supplying the assigned higher level address, for instance, the IP address in case of IP.
The IP protocol defines a particular set of control messages, known as the Internet Control Message Protocol, (ICMN). ICMP's functions are an essential part of IP. All hosts and routers must be able to generate ICMN messages as well as process the ICMN messages that they receive. For instance, ICMP provides the m
Baransky Yurij Andrij
Franke Hubertus
Pattnaik Pratap Chandra
Cameron Douglas W.
Dougherty Anne Vachon
Follansbee John A.
LandOfFree
Method and apparatus for asset tracking of network attached... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for asset tracking of network attached..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for asset tracking of network attached... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3013466