Lucent Technologies v. Newbridge Networks Corp.

• Decision Date: 21 September 2001
• Docket Number: No. 97-347-JJF.,97-347-JJF.
• Citation: 168 F.Supp.2d 181
• Parties: LUCENT TECHNOLOGIES, INC. Plaintiff, v. NEWBRIDGE NETWORKS CORP. and Newbridge Networks, Inc. Defendants.
• Court: U.S. District Court — District of Delaware

168 F.Supp.2d 181

LUCENT TECHNOLOGIES, INC. Plaintiff, v. NEWBRIDGE NETWORKS CORP. and Newbridge Networks, Inc. Defendants.

No. 97-347-JJF.

United States District Court, D. Delaware.

September 21, 2001.

COPYRIGHT MATERIAL OMITTED

COPYRIGHT MATERIAL OMITTED

COPYRIGHT MATERIAL OMITTED

COPYRIGHT MATERIAL OMITTED

COPYRIGHT MATERIAL OMITTED

Josy W. Ingersoll, John W. Shaw, Young Conaway Stargatt & Taylor LLP, Wilmington, DE, Kirkland & Ellis, New York, NY (John M. Desmarais, Gregory S. Arovas, Robert A. Appleby, Henry G. Sawtelle, James T. Bailey, of counsel), for Plaintiff Lucent Technologies, Inc.

Arthur G. Connolly, Jr., Connolly Bove Lodge & Hutz, Wilmington DE, Morgan Lewis & Bockius LLP, Philadelphia, PA (Thomas B. Kenworthy, William P. Quinn Jr., Clifford J. Peterson, of counsel), Stephen G. Rudisill, Janet M. Garetto, Russell J. Genet, Jenkens & Gilchrist PC, Chicago, IL, for Defendants Newbridge Networks Corporation and Newbridge Networks, Inc.

OPINION

FARNAN, District Judge.

Pending before the Court in this patent infringement action is a Motion For Judgment As A Matter Of Law¹ (D.I. 616) filed by Plaintiff, Lucent Technologies, Inc. ("Lucent") and a Renewed Motion For Judgment As A Matter Of Law At The Close Of The Evidence (D.I.618) and a Motion For A New Trial On Certain Issues (D.I.617) filed by Defendants, Newbridge Networks Corporation and Newbridge Networks, Inc. (collectively "Newbridge").² By its Motion For Judgment As A Matter Of Law, Lucent requests the Court to enter a judgment of infringement against Newbridge on Claims 1 and 4 of U.S. Patent No. 4,750,136.

By its Renewed Motion For Judgment As A Matter Of Law, Newbridge requests the Court to enter a judgment of non-infringement as a matter of law with regard to all five of the patents-in-suit. Specifically, Newbridge contends that Lucent failed to establish infringement by a preponderance of the evidence, or in the alternative, that several defenses preclude a finding of infringement against New-bridge. Alternatively, Newbridge seeks a new trial on the questions of infringement and damages, as well as on certain of its defenses, including the validity of the patents and the existence of an implied license to practice the patents. (D.I.617).

For the reasons set forth below, the Court will deny Lucent's Motion For Judgment As A Matter Of Law, grant in part and deny in part Newbridge's Renewed Motion For Judgment As A Matter Of Law, and deny Newbridge's Motion For A New Trial On Certain Issues.

BACKGROUND

I. Procedural Background

Plaintiff Lucent Technologies, Inc. ("Lucent") filed the instant action against Defendants Newbridge Networks Corporation and Newbridge Networks, Inc. (collectively "Newbridge") on June 24, 1997 (D.I.1), initially alleging infringement of four patents: (1) U.S. Patent No. 4,769,810 (the "Eckberg '810 patent"); (2) U.S. Patent No. 4,769,811 (the "Eckberg '811 patent"); (3) U.S. Patent No. 4,979,174 (the "Cheng '174 Patent"); and (4) U.S. Patent No. 4,437,087 (the "Petr '087 Patent"). On January 13, 1998, Lucent amended its Complaint to add a claim of infringement under a fifth patent, U.S. Patent No. 4,750,136 (the "Arpin '136 Patent"). (D.I.25, 30).

A three week jury trial was held on all issues presented by the parties, except certain equitable defenses. The jury returned a verdict in favor of Lucent on all issues, except infringement of Claims 1 and 4 of the Arpin patent. Following the trial, the parties filed a stipulation resolving the outstanding equitable issues. (D.I. 683). Accordingly, the only issues remaining for the Court's consideration are those raised by the parties' post-trial motions.

II. Technological Background

The patents at issue in this action relate generally to data networking. (D.I.551).

A. Eckberg '810 Patent and Eckberg '811 Patent

The Eckberg '810 and '811 patents disclose methods and apparatus for controlling congestion in a packet-switching network. In an end station of a packet-switched network, a data message is broken into shorter pieces called packets, which are then individually transmitted by the end station into the network. Each packet includes a header. The header contains information identifying the destination of the packet, or in some implementations, the connection to which the patent belongs. The nodes of the packet-switched network use the information in the header to route the packet through the network to the intended destination end station. At the destination end station, the packets are reassembled into the original message. (D.I. 551 at 14-15.)

When several users attempt to utilize a network at the same time, congestion can occur. Congestion occurs when a network component, or a portion of a network component, is required to process more packets than it can handle. In addition to long delays in data transmission, congestion can also result in packet loss. As described in the Eckberg patents:

A principal area of packet congestion is in buffers, or queues, in each node, particularly where the buffers become unavailable to store incoming packets. Yet the buffer requirement is closely related to the utilization of processor real time and/or link bandwidth. When the processor real time is exhausting, or when the link bandwidth is not sufficient to handle the packet traffic, queues within the switching node will build up causing a long delay. Finally packet buffers will be exhausted, resulting in the dropping of packets.

('810 patent, col. 2, 11. 26-35; '811 patent, col. 2, 11. 25-34). Congestion and the resulting loss of packets, can be prevented and controlled by allocating bandwidth to users' connections and permitting new connections only when bandwidth is available. Bandwith is the amount of data that can be sent per unit of time. To allocate bandwith appropriately in this manner, the network must provide a means for users to select their bandwith needs. Once the network understands the user's needs, the network creates the appropriate parameters for the user and enforces those parameters for the user. The agreement between the user and the network concerning the user's bandwith needs is called a traffic contract. Once a traffic contract is established between the user and the network, packets that are transmitted at a rate in excess of the agreed parameters are automatically discarded at the network's entry point.

Because traffic on the network varies at times, the automatic discarding of packets that exceed a user's allocation is not efficient. Specifically, there may be times that the network could have handled the excess traffic instead of dropping those excess packets. The Eckberg '810 and '811 patents teach a traffic enforcement mechanism that allows excess packets to be marked, instead of automatically discarded. By marking excess patents the network can decide whether to discard marked excess packets during periods of congestion, or to allow those marked packets to proceed during periods of relative quiet. Generally speaking, the claims of the '810 patent are directed toward marking excess packets, and the claims of the '811 patent are directed toward dropping the marked packets during periods of network congestion. (D.I. 551 at 16-18.)

B. Cheng '174 Patent

The Cheng '174 Patent teaches a method of handling random and bursty errors in a data transmission system. A random error is typically caused by noise and generally results in a single bit error. A bursty error is usually more serious and results in multiple bit errors. When a random or bursty error occurs in the header of a packet, that packet may be delivered to an incorrect destination address. (Cheng, col.1, 11.11-23).

Although the prior art in this field taught various techniques for dealing with these types of error, the prior art was not adequate in packet-switching systems involving high transmission speeds. High transmission speeds often limited the effectiveness of the prior art's error correction codes, because only a few bytes of error correction coding could be used if the speed was to be maintained.. (Cheng, col.1, 11.24 — 46). By using only small amounts of error correction coding, bursty errors could not be quickly and accurately detected.

To resolve the problems associated with the prior art, the Cheng Patent teaches a method to detect both random and bursty errors using only a small amount of error correction coding. The patent teaches a switch between two states: (1) an error correction circuit state (ECC); and (2) an error detection circuit state (EDC). The ECC state is used during normal conditions. The EDC state is used when an error is detected on the assumption that bursty conditions have arisen. If the assumption proves true and bursty conditions have arisen, the many errored packets that result from bursty conditions will be detected and discarded, thereby preventing their delivery to the wrong user. However, if the assumption is incorrect and bursty conditions have not arisen, it is unlikely that the next packet processed will have any errors. Accordingly, the system will switch back to the normal ECC state to perform random error corrections, rather than bursty error corrections. (D.I. 551 at 36).

C. The Petr '087 Patent

The Petr '087 Patent relates to speech compression. Speech signals are typically transmitted in digital telephone systems via pulse code modulation (PCM). PCM is a standardized technology whereby analog signals are sampled at 8000 times per second and converted into pulses. These pulses are each represented by an 8-bit code. The result is a transmission rate equal to 64 kbits/sec. (D.I. 551 at 47.)

Adaptive Differential Pulse Code Modulation (ADPCM) is a method for compressing PCM signals, thereby reducing bandwith and increasing efficiency. The Petr patent teaches an improved form of ADPCM. (D.I. 551 at 47.)

A typical ADPCM coder and decoder consists of a quantizer and a predictor, either or both of which are capable of changing...