Peter Mendham - PowerPoint PPT Presentation

About This Presentation
Title:

Peter Mendham

Description:

Title: Enabling Interoperability and Rapid Integration with the SpaceWire-PnP Protocol Subject: Introduction to SpaceWire Author: Peter Mendham Last modified by – PowerPoint PPT presentation

Number of Views:56
Avg rating:3.0/5.0
Slides: 50
Provided by: PeterM247
Learn more at: https://cwe.ccsds.org
Category:
Tags: mendham | peter

less

Transcript and Presenter's Notes

Title: Peter Mendham


1
SOIS Plug-and-Play Use Cases and Requirements
  • Peter Mendham

30 April 2016
2
Agenda
  • Introduction and background
  • Operational scenarios
  • Scenario use cases
  • Synthesising a use case
  • Service level approaches and requirements
  • Subnetwork level requirements
  • Subnetwork support requirements
  • Recommendations

3
Introduction
  • Asked to review plug-and-play
  • Use cases
  • Definition
  • Scope
  • Requirements
  • Starting point was the concept paper
  • Plus any other published material available

4
Approach
  • Back to basics
  • Largely ignore the SOIS architecture
  • Intend to review the architecture as an outcome
  • Scenarios to use cases to requirements to a
    better defined purpose and scope

5
Definitions of Plug-and-Play
  • Generic definition (various sources)
  • SOIS definition (concept paper)

Plug-and-play encompasses the characteristics of
an interface or device specification to
facilitate the discovery of a hardware component
in the system and the automatic configuration of
that component such that it may be used without
user intervention.
Plug-and-play encompasses the mechanisms
necessary to establish communication services
between two data systems in a spacecrafts
onboard (sub-)network, without requiring
reconfiguration or manual installation of device
drivers by any user (higher-level service or OBSW
application).
6
Scenarios
  • Rapid spacecraft development
  • Automated integration and test
  • FDIR assistance
  • Ground segment access

7
Rapid Spacecraft Development (1)
  • Many drivers to reduce spacecraft development
    time
  • Commercial drivers lt 2 years
  • Military drivers (e.g. ORS) lt 1 week
  • Reduce development time through
  • Standardised components
  • Standardised interfaces
  • Incremental qualification/VV (cf. IMA)
  • Use of in-house/commercial off the shelf
  • Rapid integration

8
Rapid Spacecraft Development (2)
  • Plug-and-play enablers
  • Removal of dependence on subnetwork configuration
  • Reuse of applications independently from devices
  • Reuse of devices independently from applications
  • Increase portability of applications
  • Beneficiaries
  • System integrators (e.g. primes)
  • Customers (e.g. agencies)
  • Device manufacturers
  • Permits/promotes inter-agency cooperation

9
Automated Integration and Test (1)
  • During integration a number of configurations are
    necessary for
  • Spacecraft
  • EGSE
  • Each configuration must be validated
  • Use plug-and-play to permit
  • Discovery of devices
  • Automated configuration of services/applications
  • Automated configuration of EGSE and other test
    equipment

10
Automated Integration and Test (2)
  • Plug-and-play enablers
  • Removal of dependence on subnetwork configuration
  • Configuration of communications independent to
    applications
  • Validate applications independent to
    communications
  • Validate devices independently from applications
    etc.
  • Beneficiaries
  • System integrators (e.g. primes)
  • Customers (e.g. agencies)
  • Device and EGSE manufacturers
  • Permits/promotes inter-agency cooperation

11
FDIR Assistance (1)
  • Fault detection
  • Detect the disappearance of subnetwork devices
  • Might indicate a fault
  • Isolation
  • Use subnetwork features to isolate device
  • Reconfigure routing/fail over bus
  • Recovery
  • Identify replacement device(s)
  • Reconfigure subnetwork as necessary

12
FDIR Assistance (2)
  • Plug-and-play enablers
  • Device discovery (addition/removal notification)
  • Device isolation (depending on subnetwork)
  • Transparent subnetwork reconfiguration for
    identical redundant devices
  • Beneficiaries
  • Software developers
  • System integrators (e.g. primes)
  • Customers (e.g. agencies)
  • Spacecraft operators (may see more reliable
    spacecraft)

13
Ground Segment Access (1)
  • Automated adaptation of ground segment interface
    to accommodate
  • Different spacecraft configurations
  • Different devices
  • Adapt telemetry expected
  • Format and content
  • Adapt telecommands issued
  • Types available and format

14
Ground Segment Access (2)
  • Plug-and-play enablers
  • Device discovery
  • Provide services independently from subnetwork
    configuration
  • Service discovery by ground segment applications
  • Beneficiaries
  • Operational stakeholders
  • Original space and ground segment customer(s)
    (e.g. agencies)
  • System integrators (e.g. primes)
  • Space and ground segment developers

15
Use Cases
  • Use case for each scenario
  • Presented as a process
  • Or multiple processes
  • Will eventually be drawn together into a
    synthesised use case

16
Rapid Spacecraft Development Use Case (1)
  • Use case for iterative/incremental development
    and qualification/VV
  • Use case for integration of devices into
    spacecraft
  • Device driven process for subnetwork issues
  • Application driven process for service issues

17
Rapid Spacecraft Development Use Case (2)
  1. Device is physically and logically integrated
    onto spacecraft network
  2. Subnetwork plug-and-play features discover the
    device
  3. Subnetwork plug-and-play network management
    carries out network configuration and
    configuration of subnetwork-related device
    features
  4. Subnetwork plug-and-play network management
    carries out configuration of subnetwork services
  5. Subnetwork plug-and-play assists in the discovery
    of services/capabilities of device
  6. Capabilities are exposed through to (potentially
    standardised) applications via a suitably
    standardised interface

18
Rapid Spacecraft Development Use Case (3)
  1. The application is loaded
  2. The application queries the service interface for
    the availability of a specific device or device
    type
  3. If the device is available the application
    obtains the service interface and binds to it,
    application features relating to the service are
    then enabled
  4. If the device is not available the application
    either waits and then returns to step 2
    (re-query), or disables the features relating to
    the service

19
Automated Integration and Test Use Case (1)
  • Many similarities to previous scenario
  • Two parts
  • Plug-and-play on spacecraft to adapt to
    integrated devices
  • Plug-and-play of EGSE attached to spacecraft for
    integration and test

20
Automated Integration and Test Use Case (2)
  1. Device is physically and logically integrated
    onto spacecraft network
  2. Subnetwork plug-and-play features discover the
    device
  3. Subnetwork plug-and-play network management
    carries out network configuration and
    configuration of subnetwork-related device
    features
  4. Subnetwork plug-and-play network management
    carries out configuration of subnetwork services
  5. Subnetwork plug-and-play assists in the discovery
    of services/capabilities of device
  6. Capabilities are exposed through to (potentially
    standardised) applications via a suitably
    standardised interface
  7. Applications registered to use the available
    service interfaces are either loaded or have the
    appropriate service-related functions enabled (if
    they are already loaded)

21
Automated Integration and Test Use Case (3)
  1. EGSE is physically and logically integrated onto
    spacecraft network
  2. EGSE subnetwork plug-and-play features discover
    the device
  3. Subnetwork plug-and-play network management
    detects current network configuration and
    configuration of subnetwork-related device
    features and configures EGSE subnetwork services
    accordingly
  4. Subnetwork plug-and-play assists in the discovery
    of services/capabilities of device
  5. Capabilities are exposed through to (potentially
    standardised) EGSE applications via a suitably
    standardised interface
  6. Test and/or VV applications registered to use
    the available device service interfaces are
    either loaded or have the appropriate
    service-related functions enabled (if they are
    already loaded)

22
FDIR Assistance Use Case (1)
  • FDIR process utilises plug-and-play at four
    stages
  • Initial integration
  • Fault detection
  • Fault Isolation
  • Recovery
  • Additionally, plug-and-play may be used to
    accommodate non-duplicate redundancy
  • Such as 3 of 4 arrangements e.g. tetrahedral gyro
    arrangements

23
FDIR Assistance Use Case (2)
  • Initial Integration
  • Subnetwork plug-and-play features discover both
    prime and redundant devices
  • Subnetwork plug-and-play network management
    carries out network configuration and
    configuration of subnetwork-related device
    features
  • Subnetwork plug-and-play network management
    carries out configuration of subnetwork services
  • Subnetwork plug-and-play assists in the discovery
    of services/capabilities of device
  • Capabilities of both devices are exposed through
    to (potentially standardised) applications via a
    suitably standardised interface

24
FDIR Assistance Use Case (3)
  • Fault Detection
  • The prime device ceases to be either physically
    or logically present on the subnetwork
  • Network discovery determines the absence of the
    device either through active polling or passive
    notification
  • Optionally, applications are notified of the
    absence of the device (and potential failure)
    and/or the lack of availability of particular
    services

25
FDIR Assistance Use Case (4)
  • Fault Isolation
  • Subnetwork plug-and-play network management
    reconfigures the subnetwork to isolate the failed
    prime device (if possible)

26
FDIR Assistance Use Case (5)
  • Recovery
  • Subnetwork plug-and-play network management
    carries out network configuration and
    configuration of subnetwork-related device
    features for the redundant device
  • Subnetwork plug-and-play network management
    carries out configuration of subnetwork services
    to accommodate the redundant device

27
FDIR Assistance Use Case (6)
  • Service Level Adaptation
  • Subnetwork plug-and-play assists in the discovery
    of services/capabilities of the remaining
    device(s)
  • A uniform service interface is presented to a
    (potentially standardised) application though
    synthesis of the services provided by the
    available devices
  • Applications query the service interface to
    determine the availability of replacement
    services for the ones lost
  • If suitable services are available the
    application obtains the service interface and
    binds to it, application features relating to the
    service are then enabled (this may require
    adapting the service interface e.g. type
    conversions, frame rotations)
  • If suitable services are not available the
    application either waits and then returns to step
    2 (re-query), or disables the features relating
    to the services

28
Ground Segment Access Use Case (1)
  • Splits into two parts
  • Onboard integration and discovery of devices
  • Access by ground segment applications
  • First part is the same as before
  • Process of second part depends access
  • Device oriented
  • Service oriented

29
Ground Segment Access Use Case (2)
  • Device Oriented
  • The ground segment application queries the
    spacecraft service interface for the availability
    of a specific device or device type
  • If the device is available the application
    obtains the service interface and binds to it,
    application features relating to the service are
    then enabled
  • If the device is not available the application
    either waits and then returns to step 2
    (re-query), or disables the features relating to
    the service

30
Ground Segment Access Use Case (3)
  • Service Oriented
  • Ground segment applications query the service
    interface to the spacecraft to determine the
    availability of suitable services
  • If suitable services are available the
    application obtains the service interface and
    binds to it, application features relating to the
    service are then enabled (this may require
    adapting the service interface e.g. type
    conversions)
  • If suitable services are not available the
    application either waits and then returns to step
    2 (re-query), or disables the features relating
    to the services

31
Synthesising a Use Case
  • Separate levels
  • Subnetwork level
  • Service level (device capabilities)
  • Subnetwork level is relatively easy
  • Service level splits into three approaches
  • Device-driven, device-bound
  • Application-driven, device-bound
  • Application-driven, service-bound

32
Subnetwork Level Synthesised Use Case
  1. A change is made to the devices physically and/or
    logically integrated onto spacecraft subnetwork
  2. Subnetwork plug-and-play features discover the
    addition or removal of a devices
  3. Subnetwork plug-and-play network management
    carries out network configuration and
    configuration of subnetwork-related device
    features
  4. Subnetwork plug-and-play network management
    carries out configuration of subnetwork services
  5. Subnetwork plug-and-play assists in the discovery
    of services/capabilities of device

33
Device-Driven, Device-Bound Use Case
  1. The subnetwork indicates the availability of a
    device or device class, potentially via a higher
    level service
  2. Applications registered to use the available
    device service interfaces are either loaded or
    have the appropriate service-related functions
    enabled (if they are already loaded), or are
    simply informed

34
Application-Driven, Device Bound Use Case
  1. The application queries the service interface for
    the availability of a specific device or device
    type
  2. If the device is available the application
    obtains the service interface and binds to it,
    application features relating to the service are
    then enabled
  3. If the device is not available the application
    either waits and then returns to step 1
    (re-query), or disables the features relating to
    the service

35
Application-Driven, Service-Bound
  1. Applications query the service interface to
    determine the availability of suitable services
    on the basis of the service elements they
    provide
  2. If suitable services are available the
    application obtains the service interface and
    binds to it, application features relating to the
    service are then enabled (this may require
    adapting the service interface e.g. type
    conversions)
  3. If suitable services are not available the
    application either waits and then returns to step
    1 (re-query), or disables the features relating
    to the services

36
Where Does the Device Interface Come From?
  • Something has to know how to talk to the device
    (i.e. a device driver)
  • Independent concern to how interface is exposed
  • Interface can be
  • Manually written or auto-coded
  • Generated online or offline
  • Making use of an electronic data sheet

37
Manually Generated Device Interface
Application
Subnetwork
Device
Load/Enable
Instantiated I/F
VID/PID Class ID Instance ID
Device I/F
OBC
Manual Coding
EDS
Ontology
38
Auto-Coded Device Interface
Application
Subnetwork
Device
Load/Enable
Instantiated I/F
VID/PID Class ID Instance ID
Device I/F
OBC
Auto-coding
EDS
Primitive Ops
Ontology
Manual Coding
39
Online Generated Device Interface
Application
Subnetwork
Device
Generation
Generated I/F
VID/PID Class ID Instance ID
Primitive Ops
OBC
EDS
Manual Coding
Ontology
40
Device Interface Generation Comments
  • Similarities
  • All can/must use an EDS
  • All utilise subnetwork level plug-and-play
  • Importantly
  • All are completely independent of subnetwork
  • Except for EDS provision
  • All are completely independent of service
    interface approach

41
Service Level Requirements
  • Need device enumeration
  • Need service virtualisation
  • Service interface depends on approach
  • Virtual device classes with standardised
    interfaces
  • Could permit device-driven application
    loading/enabling
  • Service interface permitting queries on service
    elements
  • e.g. I need a service which gives me the
    temperature of this part of the spacecraft

42
Subnetwork Level Requirements
  • Device discovery
  • Management of subnetwork
  • Subnetwork-specific features of devices
  • Subnetwork resources
  • Management of subnetwork addressing
  • Reconfiguration of other subnetwork services
    appropriately

43
Subnetwork Requirements
  • Network discovery
  • Device discovery (polled/notified)
  • Topology discovery
  • Device identification
  • Unique identification
  • Type/class
  • Configuration of subnetwork-specific device
    features
  • e.g. address, link speed
  • Configuration of subnetwork resources
  • e.g. bus controllers, routers
  • Sourcing of electronic data sheet

44
Subnetwork Network Management
  • Issues of mechanism
  • How addresses are set
  • Subnetwork specific
  • Independent of mission
  • Issues of policy
  • How addresses should be assigned
  • Subnetwork specific
  • Mission or mission class specific
  • Mechanism and Policy should be separated

45
Subnetwork Plug-and-Play Architecture
Applications
Device Enumeration Service
Device Virtualisation Service
Device Access Service
Network Management Service
Device Discovery Service
Memory Access Service
Packet Service
Physical Layer
46
Subnetwork Plug-and-Play Architecture
Applications
Device Enumeration Service
Device Virtualisation Service
Device Access Service
Network Management Service
Policy
Defined Interface?
Network Management Service
Configuration
Device Discovery Service
Memory Access Service
Packet Service
Physical Layer
47
Plug-and-Play Scope
  • Subnetwork and service levels are separable
  • Subnetwork level scope
  • Device discovery
  • Device identification
  • Subnetwork configuration and management
  • Service level scope
  • Service discovery
  • Standardised virtual interface
  • Query-able interface?
  • Application binding

48
Conclusions and Discussion Points (1)
  • Plug-and-play is highly beneficial
  • It would seriously hinder the progress of
    platforms using SOIS not to have plug-and-play
  • Requirements for subnetwork level are easily
    understood
  • Requirements placed on subnetwork technologies
    are easily understood
  • Scope of plug-and-play in the subnetwork context
    is easily definable and constrained
  • A couple of minor changes necessary to subnetwork
    architecture
  • Policy should be separated from mechanism
  • Subnetwork is separable from service level

49
Conclusions and Discussion Points (2)
  • A standard format for electronic data sheets
    would be useful to help tie subnetwork and
    service levels
  • Service level plug-and-play is less constrained
  • Different approaches possible
  • Based on standardised device classes known a
    priori
  • Based on a query-able service interface to permit
    service inspection
  • Should both be supported?
  • Should subnetwork and service level plug-and-play
    be separated in the context of SOIS?
Write a Comment
User Comments (0)
About PowerShow.com