Netscout Systems, Inc.

Rajeev Nadkarni

Anil K Singhal

Narendra Byrapuram

Bruce A Kelley, Jr

Abhishek Saraswati

Tauras Liubinskas

John Peter Curtin

Ashwani Singhal

Sandeep Prasad

Erdem Uysal

Mahesh Srinivasagowda

Anthony Peter Joch

Gregory Mayo

Brandon Wayne Bass

Bruce Sweet

Amogh N Joshi

Mahesh Rajmane

Bruce Kelley

Prashant Pharande

Ah Htwe Yi

Nilesh Tayade

Claudio Guerrini

Saraswati Saraswati

A system having a network monitor communicating with U and C probes monitoring SGW-U and SGW-Cs. The U probes buffer captured user plane packets from which they determine new sessions and send new session data to the network monitor, the new session data including an IP address pair of the corresponding U probe and SGW-U. The C probes inspect captured control plane packets, determine control plane packets having a same IP address as an SGW-U in the new session data, and send associated control data to a particular U probe having the IP address paired with the IP address of the SGW-U, wherein the particular U probe correlates the control data with buffered user plane packets and generates associated monitoring data.

Matching user and control plane data in a network system using control and user plane separation

A system and method for monitoring one or more Mobility Management Entities (MMEs) with a plurality of scalable network probe devices arranged in a cluster format. A ciphered packet is received from one or more MMEs at a packet switching device. The packet switching device in turn sends all the ciphered packets to each of the plurality of clustered probes. Each of the network probes then in turn deciphers the packets received from the MME and extracts metadata from the deciphered packet to identify subscriber session information contained in the received packet. Each of the network probes then selectively retains deciphered packet information relating to a subscriber session and/or other prescribed criteria designated for that particular network probe and discards the remaining deciphered packet or portions of the packet so as to balance the load amongst the plurality of probes based upon prescribed load balancing criteria. KPI and other session related data is generated in a network probe associated with a subscriber session from the retained deciphered packet information. Subscriber session related data from each clustered network probe is then aggregated with at least one monitoring device operably coupled to the clustered probes such that a user of the monitoring device is provided with the perception that the monitoring device is coupled to a single probe.

System and method for load balancing of network packets received from a MME with smart filtering

A computer implemented method for determining the identity of an Over-the Top (OTT) application or service being accessed over the Internet from a HTTP, HTTPS or QUIC connection request received in a network monitoring device. Determine if one or more entries are present in the received connection request have an IP address that matches a known server IP address. A determination is then made as to whether if the received connection request is one of a HTTP, HTTPS or QUIC connection request, and if this cannot be determined than determine if a subject field in the received connection request is available. And determine if a candidate domain name is available from IP cache created from one or more of the above steps if a subject field is not available in the received connection request. Identify and categorize OTT applications associated with the received connection request if it is determined: the connection is either a HTTP, HTTPS or QUIC connection type; a subject field is available; or a candidate domain name is available utilizing a lookup table that is periodically updated with new OTT applications.

System and method for identifying OTT applications and services

A method is disclosed. In the method, a data generation process can continuously generate data in real time. The data generation process can store the data into discrete data blocks. An analyzer process can run analytical queries on the data from the data blocks. After the analytics is complete for different data blocks, data can be removed from the respective data blocks. The empty data blocks can be returned back to the generation process for reuse. The data blocks can be shared resources between the generation and the analyzer processes. The data can be stored in a directly queryable format. Though at any given time a given analytical query can run on a single data block, the analyzer process can preserve certain important records from that data block to be used while analyzing subsequent data blocks at a later time.

Architecture for a multi-resolution, in-flight and programmable analytics platform

A cluster load balancer can be coupled to a network for monitoring network traffic. The cluster load balancer can include one or more computing devices. The one or more computing devices can be configured to receive a data packet, the data packet comprising a device identifier of a first device connected to a network for a communication session, wherein the data packet is a user plane data packet or a control plane data packet; generate a probe identification based on the device identifier of the first device in the data packet; add the probe identification to the data packet; and transmit the data packet with the probe identification to a second device in communication with a plurality of network probes. The second device can be configured to forward the data packet to a network probe of the plurality of network probes based on the probe identification in the data packet.

Smart cluster load balancing

Method for monitoring network performance in a telecommunication network coupled with a plurality of Virtual Machines (VM) arranged in a cluster format is disclosed. A packet is received at a VM smart cluster device. Metadata is extracted from the packet. The packet can be distributed to one of the plurality of VMs. Key performance indicator (KPI) session related data associated with a subscriber in one of the plurality of VMs that receives the distributed packet can be generated.

System and method for monitoring network performance

A system and method for monitoring one or more Mobility Management Entities (MMEs) with a plurality of scalable Virtual Machines (VM)/probes arranged in a cluster format. A ciphered packet is received from a MME at a smart cluster device/probe whereby data is aggregated from the individual clustered VMs/probes for distribution to a monitoring device. The smart cluster device/probe is preferably configured to decipher the ciphered packet received from the MME and extract metadata from the deciphered packet to identify subscriber information for the received packet. The deciphered packet is then distributed to one of the plurality of clustered probes to balance the load amongst the plurality of clustered probes. The balancing of loads is based upon prescribed load balancing criteria such that each packet received for an identified subscriber is sent to a same probe such that load balancing is performed on a per subscriber basis and/or with other state-based criteria. KPI session related data associated with a subscriber is generated in a clustered probe that receives the distributed packet from the smart cluster device/probe. Subscriber related data from each clustered probe is then aggregated with at least one monitoring device operably coupled to the clustered probes such that a user of the monitoring device is provided with the perception that the monitoring device is coupled to a single probe.

System and method for load balancing of network packets received from a MME with smart clustering

A network monitoring system that summarizes a plurality of data packets of a session into a compact session record for storage and processing. Each session record may be produced in real-time and made available during the session and/or after the termination of the session. Depending on protocols, a network monitoring system extracts different sets of information, removes redundant information from the plurality of data packets, and adds performance information to produce the session record. The network monitoring system may retrieve and process a single session record or multiple session records for the same or different protocols to determine cause of events, resolve issues in a network or evaluate network performance or conditions. The session record enables analysis in the units of session instead of individual packets. Hence, the network monitoring system can analyze events, issues or performance of the network more efficiently and effectively.

Real-time adaptive processing of network data packets for analysis

A system for selective user plane (UP) monitoring includes a service gateway (SGW) having a plurality of units. The system further includes a network packet broker (NPB) configured to receive packets including UP data from tunnels created to enable transmission of the UP packets from UE to the plurality of SGW units. The NPB is also configured to receive packets including control plane (CP) data from channels enabling transmission of the CP packets from a base transceiver station to the SGW. The system also includes a plurality of probes operatively coupled to the NPB. The probes are configured to generate first metrics associated with the received CP packets and to selectively generate second metrics associated with the received UP packets based on one or more identifiers. The NPB is configured to forward UP packets being processed by a particular SGW unit to a particular probe of the plurality of probes.

Selective user plane monitoring multiple monitoring probes when a serving gateway has multiple IP addresses

A system and method to aggregate subscriber-perspective data from live data packets of network traffic. The method includes inspecting live packets of network traffic exchanged with a plurality of subscriber end devices. Network traffic exchanged with a subscriber end device can include network traffic exchanged with a different subscriber end device or with an application server. The packet inspection is performed at a location of interception of the live packets, each subscriber end device being an end device correlated with a subscriber. The method further includes selecting, at the location, content data of the inspected packets that correspond to packets exchanged with a selected subscriber end device of the plurality of subscriber end devices, aggregating, at the location, the content data selected, wherein the content data has not been previously aggregated, and outputting, at the location, results of the aggregation.

System and method for aggregating subscriber perspective data

A system and method is provided for calculating session initiation protocol (SIP) key performance indicators (KPIs) for multimedia communication in a network accessing network packets monitored over time. The method includes accessing network packets monitored over time, detecting a SIP response message of the monitored network packets, and accounting for the detected SIP response message in a time window of a series of consecutive time windows that includes the SIP response message's associated observed time. The SIP response message has an associated observed time at which the monitored SIP message was observed during monitoring. The method further includes pairing a SIP request message expected to be paired with the detected SIP response message with the detected SIP response message and accounting for the paired SIP request message in the time window with which the detected SIP response message is accounted. The method further includes calculating a KPI for the time window using the detected SIP response message and its paired SIP request message that are accounted for in the time window.

Growing community of inventors

Chelmsford, MA, United States of America

Rajeev Nadkarni