Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Cognitive Radio and Network R&D Trial Environment


Published on

The CORE trial environment provides functions such as obtaining knowledge, making decisions and adjustments on the system under cognitive control. The trial environment provides a cognitive engine for decision making, a WWW browser user interface and CORE tools for collecting knowledge and making adjustments.

See for more information

Published in: Technology
  • Be the first to comment

Cognitive Radio and Network R&D Trial Environment

  1. 1. Cognitive Trial EnvironmentCORE Workshop 15.1.2013Marko PalolaVTT Technical Research Centre of Finland
  2. 2. 21/03/2013 2 CORE project 2011-2014Develop and deploy future CR business ecosystem techniquesDevelop a cognitive trial system for carrying out experimentsMake proof of concept trials and measurements VTT, Centria, CWC NSN, EXFO Nethawk, EB, Pehutek, PPO, Renesas, Finnish Defence Forces TEKES Trial Programme 15.1.2013 Marko Palola, VTT 2
  3. 3. 21/03/2013 3 Cognitive Radio Systems“A radio system employing technology that allows the system to obtain knowledge of its operational and geographical environment, established policies and its internal state; to dynamically and autonomously adjust its operational parameters and protocols according to its obtained knowledge in order to achieve predefined objectives; and to learn from the results obtained.” REPORT ITU-R SM.2152 15.1.2013 Marko Palola, VTT 3
  4. 4. 21/03/2013 4 Cognitive Radio SystemsEnabler for next generation,dynamic spectrum access Use or share the spectrum in annetworks opportunistic manner Best connection Power consumption Load balancing Network optimisation Resource sharing Network reconfiguration 15.1.2013 Marko Palola, VTT 4
  5. 5. 21/03/2013 5 Cognitive Cycle realisation in the CORE project 1/2 Internal stateis specified by the collected General data approach for decision making based on data Collection ofnetwork information, Resource load user QoS, user balansing location, and radio resource data. adjustments. Generalized 15.1.2013 Marko Palola, VTT implementation 5
  6. 6. 21/03/2013 6 Cognitive Cycle realisation in the CORE project 2/2Obtaining knowledge & Adjustments CORE Client tools CORE Cognitive API CORE Cognitive ManagerDecision making Data cache & history Cognitive Engine Decision algorithmTrial Access CORE Trial Environment User Interface 15.1.2013 Marko Palola, VTT 6
  7. 7. 21/03/2013 7 Trialing with CORECORE trial environment allows integrating an external system undercognitive engine control. Requires installing/development of Core Client tools for information collection and making adjustments During trial Trial customer is able to make decision rules based on collected data that change the behaviour of the external system Use and measure the external system 15.1.2013 Marko Palola, VTT 7
  8. 8. 21/03/2013 8 Obtaining Knowledge & Making changesCore Clients interface the externalsystem, which may consist of Computer, terminal, sensor Server, router, base station, switch Simulator, measurement or monitoring tool Database, file, socket Service platform Any equipment that can be monitored 15.1.2013 Marko Palola, VTT 8
  9. 9. 21/03/2013 9 Obtaining Knowledge & Making changes Core Clients are information collectorsCore Clients monitor / receive information fromexternal systemProduce data for decision making using COREdata formatData delivery as soon as information is changed Periodical status (refresh current state) 15.1.2013 Marko Palola, VTT 9
  10. 10. 21/03/2013 10 CORE data formatCORE uses the following data presentation format component.parameter = value; Examples: basestation.load = high; basestation.load=88; network.available=wlan1; network.current=centria_lte; Names and values can be freely selected. Cognitive engine follows values and changes related to user, terminal, 15.1.2013 network, server, service, and basestation components. Marko Palola, VTT 10
  11. 11. 21/03/2013 11 Event message Core clients deliver collected information to their associated CE for decision making using an event message specified by Cognitive API. Who/where collected information How long data is usable? Data Related to component.parameter = value; what? 15.1.2013 Marko Palola, VTTReferences: 4Ward, Univerself, Future Internet, Mercone… 11
  12. 12. 21/03/2013 12 Obtaining Knowledge & Making changes Core Clients configure the external system Core clients implement routine that make changes to external system. Core clients inform the cognitive engine what they can do A core client’s capability activated by the cognitive engineCORE project have developed capabilities for Handover capability Offloading, load balansing WARP radio network channel management channel occupancy information 15.1.2013 Marko Palola, VTT 12
  13. 13. 21/03/2013 13 Decision making in CORECognitive engine Rule-based decision making (if .. then ..) Decision rulesDecisions are based on Data on incoming event message Earlier received data (cache)Decision rules can be defined by the Trialcustomer. 15.1.2013 Marko Palola, VTT 13
  14. 14. 21/03/2013 14 Decision algorithmA bunch of decision rules, each rule has Requirement(s) Capability actionRequirement is tested each time new data event arrives on thedecision engineCapability action is defined by a core clientIf requirements are all true the capability action is activated -> an event is sent to the associated core clients 15.1.2013 Marko Palola, VTT 14
  15. 15. 21/03/2013 15 Decision Rule: RequirementsRequirement is a conditional clause formatted as follows component.parameter<operator>value;Component is a text containing user, terminal, network, service, basestation or free text Parameter is a text such as name, location, id, user, terminal, network, delay, throughput, .. Operator is !=, ==, >, < , inc, not Value is a number or textExample: network.downlink>21.5; 15.1.2013 network.downlink=low; Marko Palola, VTT 15
  16. 16. 21/03/2013 16 Decision Rule: Capability actionCapability can be defined by 1) Event component.action(value); 2) Event/value;component.action(value); 3) Event component.parameter=value;In 1) and 3) the capability action is sent as it is declared to the interestedparties.In 2) the cognitive engine will seek for value from the data cache and set thevalue inside the ().For example: HandoverEvent/network.available;;May result the following (depending on the network.available known by CE) HandoverEvent; 15.1.2013 Marko Palola, VTT 16
  17. 17. 21/03/2013 17 Cognitive API – event formats• Data and other control information• Information routing• Control who access what• Animation purposesReferences: 4Ward, Univerself, Future Internet 1-2, …
  18. 18. 21/03/2013 18CORE general architecture 15.1.2013 Marko Palola, VTT 18
  19. 19. 21/03/2013 19Examples of Core Clients developed in CORE projectHandover / Data offloading HandoverCoreClient NetshCoreClientQoS measurement data collection QosmetCoreClientUser location in LTE network LteCoreClientWARP radio resource management WarpCoreClient 15.1.2013 Marko Palola, VTT 19
  20. 20. 21/03/2013 20 HandoverCoreClientExternal system Birdstep SafeMove application VTT’s Mobile IP service Seamless handoverHandoverCoreClient Current network Available network Handover capability 3G/4G-WLAN-LAN Leave network 15.1.2013 Marko Palola, VTT 20
  21. 21. 21/03/2013 21 HandoverCoreClientExternal system netsh ipconfigCore Client Example information and capability definitions for handover core clients: Current network NetworkFoundEvent network.current=LocalAreaConnection; Available network NetworkFoundEvent network.available=panoulu; Handover capability HandoverEvent/network.available;; WLAN-WLAN HandoverEvent; 4G/3G-WLAN-LAN HandoverEvent handover.away(panoulu); 15.1.2013 Leave network Marko Palola, VTT 21
  22. 22. 21/03/2013 22 QoS information collection for decision makingExternal system: QosmetCoreClient:Qosmet QoS measurement system • Monitors measured QoS data for changes -> send message for decision makingQoS measured for both directionsbetween the measurement points for • Configurable limits for QoS data valuesselected data flow. -> send message for decision making 15.1.2013QoS reported every second Marko Palola, VTT 22
  23. 23. 21/03/2013 23 QoS information collection for decision makingQosmetCoreClient receives the following QoS parameters(typically in 1 seconds interval): measurement duration offered load uplink timestamp bytes received uplink packet loss packets received uplink packet loss total bytes sent downlink packet loss packets sent downlink packet loss total downlink calculated MOS uplink packet delay uplink calculated MOS downlink packet delay uplink packet jitter Example decision data delivered by QosmetCoreClient downlink packet jitter • throughput.speed=low; • throughput.value=345.22; throughput downlink 15.1.2013 Marko Palola, VTT 23
  24. 24. 21/03/2013 24 LTECoreClient – user LTE cell location LteCoreClient receives various information from EAST tool:External system: • CREATE_ACK – an LTE device has attached to the network• Centria LTE test network • HO_EVENT – an LTE device has made handover • DESTROY_ACK – an LTE device is disconnected• NetHawk East LTE tool • ENB_ASSO_INFO – an LTE base station has been started • STARTUP – the network is restarting Delivers user cell location and number of clients in a cell to decision making • UE2.location = 65311; •; 15.1.2013 • basestation.cellid=25601;basestation.clients=1; Marko Palola, VTT 24
  25. 25. 21/03/2013 25 WarpCoreClient – WARP radio NW resource managementExternal system:• CWC’s WARP radio network• WARP databaseWarpCoreClient brings WLANchannel availability information intoWARP radio network• Sets free / occupied WLAN channelsWARPs avoid reserved WLAN 15.1.2013channels Marko Palola, VTT 25
  26. 26. 21/03/2013 26 Summary of CORE Trial Environment featuresCORE Cognitive engine CORE Client-side tools Integrate & control new systems easily Windows 7-based network detection, handover Stand-alone and distributed decision MIP network detection, seamless handovers making Qosmet 2-way network QoS information collector – delay, Cognitive API for data collection delay jitter, packet loss, throughput, General QoE (1-5) Browser-based user interface LTE user location monitoring – Cell id, IMSI, IP address Decision rule editor MIH Network Information data source (on development) Monitor collected data Networks, basestations, basestation status Visualisation of data WARP Cognitive radio network resource management Configuration of client-side tools, policies, (on development) access, data types (on development) WLAN channel management over network Cognitive Engine hosting at VTT CNL LTE network resource management (2013) 15.1.2013 Marko Palola, VTT 26
  27. 27. 21/03/2013 27 VTT - 70 years oftechnology for business and society 15.1.2013 Marko Palola, VTT 27
  28. 28. 21/03/2013 28DEMONSTRATIONS 15.1.2013 Marko Palola, VTT 28
  29. 29. 21/03/2013 29CORE Trial Environment Demo WLAN/LTE offloading 15.1.2013 Marko Palola, VTT 29
  30. 30. 21/03/2013 30 CORE Trial Environment Demo Centria IPTV priority (LTE-WLAN offloading) Trial users Qosmet Qosmet QosmetCoreClient QosmetCoreClient HandoverCoreClient HandoverCoreClient MIP Client MIP Client Puuhkala Ratakatu S1+X2 S1+X2 S1+X2 Ylivieska VTT PPO S1+X2 CORE Trial Environment Alavieska S1+X2 iPro Cognitive Engine Kaisaniemi EXFO EAST LTE Visualization M5 Analyzer MME-SGW- Simulator BTS Site Manager LTECoreScript CENTRIA Cognitive API Cognitive Manager Cognitive Engine LTECoreClientWi-Fi2 Qosmet QosmetCoreClient HandoverCoreClient MIP Client EAST WWW visualization Wi-Fi1 UI 15.1.2013 Marko Palola, VTT 30
  31. 31. 21/03/2013 31 15.1.2013 Marko Palola, VTT 31