Keynote Paper

Open Architecture Applied to Next-Generation Weapons

Leo J Rosea, Jonathan Shavera, Quinn Youngb, Jacob Christensenb

a
Air Force Research Laboratory, 101 W Eglin Blvd, Suite 304, Eglin AFB, FL, USA 32542;
b
Space Dynamics Laboratory, Utah State University Research Foundation, 1695 North Research
Park Way, North Logan, UT, USA, 84341
ABSTRACT
The Air Force Research Laboratory (AFRL) has postulated a new weapons concept known as Flexible Weapons to
define and develop technologies addressing a number of challenges. Initial studies on capability attributes of this
concept have been conducted and AFRL plans to continue systems engineering studies to quantify metrics against which
the value of capabilities can be assessed. An important aspect of Flexible Weapons is having a modular plug-n-play
hardware and software solution, supported by an Open Architecture and Universal Armament Interface (UAI) common
interfaces. The modular aspect of Flexible Weapons is a means to successfully achieving interoperability and
composability at the weapon level. Interoperability allows for vendor competition, timely technology refresh, and
avoids costs by ensuring standard interfaces widely supported in industry, rather than an interface unique to a particular
vendor. Composability provides for the means to arrange an open end set of useful weapon systems configurations. The
openness of Flexible Weapons is important because it broadens the set of computing technologies, software updates, and
other technologies to be introduced into the weapon system, providing the warfighter with new capabilities at lower
costs across the life cycle. One of the most critical steps in establishing a Modular, Open Systems Architecture (MOSA)
for weapons is the validation of compliance with the standard.
Keywords: MOSA, Open Architecture, Flexible Weapons, Interoperability, Composability

1. INTRODUCTION
An important aspect of AFRLs Flexible Weapons initiative is the ability to create a family of weapons using common
functions and components. While there may be a penalty paid in individual unit costs or performance compared to a
closed, tightly coupled / integrated weapon, it is believed that the overall cost of ownership of the Open Architecture
based weapon system will be greatly reduced, considering the ease of component interchanges, technology refresh rates,
product improvements, hardware testing, and software updates. In order to achieve this desired end-state, AFRL needs
to create a Plug-n-Play environment, based on modularity, in a comprehensive system of systems through the
establishment of an Open Architecture.
Modularity requires encapsulating functionality within a physical unit with clearly defined interfaces. This can present a
challenge when functions in different regions may be tightly coupled. A sensor, for instance, may be separate from the
processing element which needs to be aware of the sensors characteristics in order to accurately transform the sensors
raw data into measurement information. One form of modularity would allow the definition of a distributed subsystem,
in which functionality is distributed across physical elements with inter-module communication described in generic
terms, with the understanding that functionality at either end of the communication paths would be responsible for
transforming data into a usable form by other subsystems. This concept enables the definition of a flexible system
architecture while protecting proprietary data that may need to be exchanged between nodes in a larger system.

2. OPEN ARCHITECTURE
Modular designs encapsulate selected functionality into separate physical units with clearly defined interfaces. What
functions are encapsulated and where physical boundaries are drawn are driven by the goals for modularity. Clearly
defined goals for modularity help design teams to understand how to evaluate different architecture options. Drivers for
modularity may include technology risks, new technology insertion, mission adaptability, maintenance, cost, and
protection of vendor intellectual property. A modular open system architecture reduces overall costs through reduced
non-recurring engineering, reduces development timelines by building on existing infrastructure, reduces the scope of
and increases automation of integration and tests, improves upgradability and maintainability through modularity, and
Open Architecture/Open Business Model Net-Centric Systems and Defense Transformation 2014,
edited by Raja Suresh, Proc. of SPIE Vol. 9096, 90960K 2014 SPIE
CCC code: 0277-786X/14/$18 doi: 10.1117/12.2055266
Proc. of SPIE Vol. 9096 90960K-1
Downloaded From: http://proceedings.spiedigitallibrary.org/ on 03/06/2015 Terms of Use: http://spiedl.org/terms

improves inter-compatibility within a system of systems. Characteristics of an open architecture include packetized,
networked data transfer which allow data to seamlessly move throughout the network, open standards that allow each
component within the system to work together, and unit and system verification / certification, allowing independent
development of components.
2.1 Architecture Overview
A system architecture for an aircraft and weapon can be depicted as in Figure 1 below. A weapon system, consisting of
a sensor, aperture, processor, payload, propulsion, and actuators must physically and logically connect to the aircraft to
make up a system of systems. In this case, the suspension and release equipment provide for the physical connection
and the stores management system provides for the logical connection.
Cockpit
Displays

J.1
Cockpit
Controls

Display
Computer

Power
Systems

T
Weapon

Stores

Aiming
Com uter

Flight Control
System

4--> Management
System
'

Engine

Management

Navigation
Computer

Radar

Inertial
Measurement
Unit

Weapon System
Sensors

Apertures

GPS Receiver

System
Suspension
and Release
Equipment

Actuators
Processors

Payload / Propulsion
Effects

Navigation
Aids

Figure 1. System Architecture for Aircraft / Weapon

2.2 Incorporating a Modular Open System Architecture

A Modular Open System Architecture (MOSA) can enable future plug-and-play weapons systems. MOSA
provides the standards and infrastructure for the capability. The vision for a plug-n-play weapon system
incorporates several features:
Weapon systems have standard interfaces to the aircraft and an electronic data sheet that includes a machine
readable (electronic data) interface control document (ICD)
The Stores Management System (SMS) is able to query all connected munitions (weapons), receive their
electronic data sheet, and automatically populate the type, capabilities, data needs, and other parameters
associated with the munition
The SMS knows from the electronic data sheets the parameters needed to define performance enveloped,
operational constraints, and display parameters
The Display Computer uses the data in the SMS to generate munition-specific cockpit displays (the munitions
electronic data sheet includes sufficient information to enable this, parameterized to maintain portability and
adaptability)
MOSA provides the common interface and the ability to self configure and self recognize, essentially automating
tasks that are done manually or semi-manually (one platform at a time) in current systems. This is shown in Figure
2.

Proc. of SPIE Vol. 9096 90960K-2

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 03/06/2015 Terms of Use: http://spiedl.org/terms

Cockpit
Displays

Display
Computer

Power
Systems
Stores

Cockpit
Controls

Flight Control
System

Management
S stem

Engine

Management

Navigation
Computer

Inertial
Measurement
Unit

Radar

System
Suspension
and Release

Equipment
Actuators

Weapon System
Sensors

Apertures

Processors

GPS Receiver

Payload / Propulsion
Effects

Navigation
Aids

Plug & Play System Architecture for Weapons System Interface adds:
Self Configuration (of aircraft, aircraft & weapon system interfaces, display info., etc.)
Self Recognition (of weapons system types, numbers, etc.)
Electronic data sheets (with specifications, options, capabilities, interfaces, etc.)
User or uploadable configuration of weapons system parameters /options, etc.

Figure 2. System Architecture and the Stores Management System

2.3 Protecting Intellectual Property Within a MOSA

One major concern in dealing with MOSA is the notion that an open system removes the ability for businesses to
maintain their competitive edge and that their intellectual property will be given away to others. The change to an
open system is better characterized as defining the interfaces, not the activity inside the functional components
within the system. In an open system, the interfaces must meet the standards as shown with the green arrows in
the diagram below, but inside the box you could have either a proprietary or non-proprietary component,
depending on what fills the need the best. Pictured in Figure 3 below, is a random division of functions that can all
work together because of the open interface, but that could be either proprietary or non-proprietary

Proc. of SPIE Vol. 9096 90960K-3

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 03/06/2015 Terms of Use: http://spiedl.org/terms

Cockpit

Display

Displays

Computer
Weapon
Aiming
Computer

Power
Systems

Stores

Flight Control

Management
S stem

System
Engine

Management

Navigation
Computer

Inertial
Measurement
Unit

Radar

System
Suspension
and Release

Weapon System
Sensors

Apertures

Equipment
Actuators
Processors

GPS Receiver

Payload /
Propulsion
Effects

Proprietary
Non -proprietary

Navigation

Open

Aids

Figure 3. System Architecture and Proprietary / Non-Proprietary Functions

Through a MOSA approach, interchangeability, reduced non-recurring engineering (NRE), and other benefits can be
realized without eliminating the intellectual property inside the individual components. This protects the competitive
edge and incentives to innovate that are important for business and Government.
2.4 Functional Decomposition

Function

Module

Function

Module

Function

Module

Function

Module

Function

Module

Modular Architecture

Integral Architecture

Functional decomposition is a method for determining logical separation of functions into groups of physical modules.
Careful functional decomposition reduces coupling between modules, and enhance the flexibility of the overall system,
as illustrated in Figure 4. For Flexible Weapons, the architecture under development will look at logical functional and
physical boundaries to separate the modules. A notional architecture, illustrated in Figure 5, shows how modules can be
organized to enable the desired composability of the system. Various interchangeable modules are available to compose
a system with the needed seeker, guidance, effect, and aerodynamics for a particular mission type.

Function

Module

Function

Module

Function

Module

Function

Module

Function

Module

Figure 4. Functional Decomposition Provides a Way to Define Physical Modules Whereby Coupling is Minimized and
Modularity is Maximized

Proc. of SPIE Vol. 9096 90960K-4

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 03/06/2015 Terms of Use: http://spiedl.org/terms

Aperture 1

Payload ;

Prop/
Glide 1

Payload ;

Prop/
Glide2

Aperture 2
Sensor 2

t
Sensor 31

Aero
Ctrl 2

Aperture 3

Prop/
Glide 3

Payload 3

Aperture 4

-- - - i
Aero
Ctrl 1

Aperture 3

Prop/
Glide2

Payload 2

Flexible Weapon
Figure 5. Flexible Weapon Notional Functional Decomposition

2.5 Architecture Development Plan

In order to successfully develop an open system design for weapons, an iterative, two-cycle approach will be
applied. The first cycle will include the first revision of architecture development, while demonstrating modularity
with bench-top hardware and preliminary software. The second cycle matures the architecture while implementing
the first revision in the second demonstration. This two-cycle approach is shown graphically in Figure 6 below.
The first demonstration provides lessons learned early enough to feed into the architecture development, and builds
momentum, while the second architecture provides a reference implementation and demonstration of the
architecture with more flight-like components in preparation for transitioning to the next phase.
Part 1

Part 2

Government and
industry SMEs inputs
Define functions .
interfaces, and
module breakout

Part 3

Part 4

Mature architecture
using industry inputs

Develop architecture
standards
Prepare for
dissemination

Develop rev FLEX

architecture
Develop use cases,

functional mapping,
module definition

Cycle 1
2014

Cycle 2
2015

Develop prototype

software using SSM - First Demonstration

of prototype using

Build test bed and

prototype with
existing hardware

test bed HW /SW

Industry builds
seeker modules and
processor module

11

Figure 6. Flexible Weapons Two-Cycle Development Plan

Proc. of SPIE Vol. 9096 90960K-5

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 03/06/2015 Terms of Use: http://spiedl.org/terms

Second demo using

industry modules
and updated
HW /SW

3. MOSA
Modular Open System Architecture (MOSA) provides Flexible Weapons the framework for incorporating desired
attributes. The building blocks of the MOSA puzzle include software modularity, hardware modularity, open standards
and interfaces, and a networked data transport mechanism. Figure 6 illustrates these building blocks and the kinds of
attributes they provide.

MOSA with net -centric component added:

Software Modularity

Physical Modularity

Hardware independent
Adapts to changes with
electronic ICDs
Standard interfaces

Expandable

Add future capability

Fully reusable module:_

Software applications
support different
missions & payloads

Networked
Decouples software from
physical location
Packetized (easy translation)
Enables security auditing
Foundation for Multi- Layered
Security (MLS)

Open
Open license standard
Full insight into workings
Improves interchangeability

Figure 7. MOSA Building Blocks

3.1 Software Development and MOSA

In following a MOSA approach, AFRL is building on the billions of dollars spent by the computer and networking
industry that has already matured the architecture and the tools. AFRL can adopt methods for software
development, data transfer, and open interfaces from the best proven practices of industries that already are familiar
with this approach, such as PCs, cell phones, etc. This implementation does not result in scraping all the software
weve already built, only changing how it works together and with the hardware so that we improve the reusability
and build on a common infrastructure.
One of the improvements in the software architecture comes from using middleware. Having the middleware in the
software stack decouples the software applications (algorithms, controls, etc.) from the hardware, as illustrated in
Figure 8. That means the software is more portable (can be used on different hardware) and doesnt have to change
every time the hardware is changed. Reuse of existing software allows more experience with individual code (find
bugs, know how it works & reduce learning curve), reducing risk, as well as reducing testing time and costs
(because the software changes less frequently).

Proc. of SPIE Vol. 9096 90960K-6

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 03/06/2015 Terms of Use: http://spiedl.org/terms

Building on billions of dollars of investment in smart architecture

Architecture Improvement:
Was:

Software
Applicatio

Is:

Software Applications
s

Layered Approach

Flight software code

Software Ser ices

Software Services

Infrastructure (common)
Changes in hardware interface

Hardware Interface
Hardware

Flight hardware

Middleware:
Decouples hardware from software
Reduces need for software changes

Not a lot more infrastructure, just a

smarter way to organize and connect
Figure 8. Smarter Software Architecture Improves Software Reuse

3.2 MOSA Improves Software Reusability

We often say that roughly 80% of software is reused from one program to another. Unfortunately, that typically means
that 80% of each software component is reused, while the remaining 20% requires changes to adapt to new hardware. In
the old paradigm, introducing new hardware for a new mission means the interfaces between hardware and software
changes (they are usually hard-coded, at least to some extent). This is costly in recreating the hardware/software
interface this is like creating Facebook and then rebuilding the internet and rewriting a web browser just to access it.
Businesses have figured out how to use the existing infrastructure on the internet and web browsers to eliminate the
interface problems.
The industry is very good at developing software the traditional way but there is a better way that will further reduce the
cost of software. Software is often the highest technical risk area and by reducing the amount of software that is written
for each system, the risk is reduced. With software reuse, there is a reduction in both the non-recurring engineering
(NRE) and the testing costs.

Proc. of SPIE Vol. 9096 90960K-7

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 03/06/2015 Terms of Use: http://spiedl.org/terms

80% software reuse / 100% re -test

Reused

Reused

Reused

Reused

Reused

w /mod

w /mod

w /mod

w /mod

w /mod

-M.

80% software reuse / 20% re -test

rer

cost

16\ Adapting

4/

to new
hardware

'-offer cost

Changes = $$ + days (implementing change and validation /verification)

We know how to build software the traditional way, but there is a better way

Figure 9. Software Reuse Saves Time and Money

3.3 Packetized Data Transfer Improves Modularity

Packetized data transfer is one way to modularize the data path, and when coupled with encapsulation, significantly
simplifies how we handle data. This is the way terrestrial data networks work like sending email from a smart phone,
a desktop machine, or a web browser: each system can send data across different kinds of networks and still work
seamlessly with each other.
Figure 10 illustrates how packetized data transfer enables a modular data path.
As a first example of a packetized data path, the data packet, A is generated from the source, or Sender, is
encapsulated when transferred from the internal network (v to w) to the first external network piece (going from w
to x). When transferred to the next external network piece (x to y) there is a different protocol and packet type, but
A is just encapsulated in the new type (B). Upon reaching the destination it is delivered as A, just like when it was
generated.
In a second example you can see the modularity: if you change the source packet type from A to D, the rest of the
network can remain the same and still get the data where it needs to go.
In a third example, changing the packet type for the w to x network from C to E type packets, the encapsulated
D packet doesnt have to change at all. It just works.
The architecture minimizes change propagation and maximizes compatibility which is why the internet works this
way, too.

Proc. of SPIE Vol. 9096 90960K-8

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 03/06/2015 Terms of Use: http://spiedl.org/terms

J
Example 1

Example 2

Example 3

Use packet encapsulation to minimize change propagation

Packet encapsulation enables modularity in network communication

Packet formats can change, other formats stay the same
Figure 10. Packet Encapsulation

The basic architecture shown in the previous illustration may look more familiar if you look at a container transportation
system, illustrated in Figure 11 below. They are the same basic architecture.
Packaging bits into sub frames and frames, then creating packets out of them is like putting books (data) into boxes (sub
frames) and boxes into pallets (frames) then pallets into a container (packet).
Once packetized, the method for delivery can be varied from trucks, to cranes, to boats, to trains, and the same packet
just rides along whatever the method. It can then be unpacked at the end and is still the same data.
To show the modularity, which allows upgradability or variability, imagine the ship is obsolete and is replaced by a
containerized aircraft. The transportation process doesnt change, and the packets dont change. The books dont care if
they are transported via boat or plane, they just get where they need to go.
Once the system is put together, the end user doesnt need to understand the complexities of the system. It just works.

Proc. of SPIE Vol. 9096 90960K-9

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 03/06/2015 Terms of Use: http://spiedl.org/terms

4*
ITOOOTOTIOTOOT
otTTOT

ro it

OT ToOT

ro to00

'CT

LL

'WNW"
011181111111

e VI

V
v

ma

=I

VCOOOTOTTOTOOT

TOOTOTOTITOTIO
TOTTOOTOTTOTOT

Figure 11. Packet Movement Through the System

3.4 MOSA Can Provide Additional Benefits

Figure 12 illustrates several additional benefits and characteristics of a MOSA-based system.
In the left side of the figure, you can see that the MOSA portions of the system can work with heritage components
(actual hardware, or networks, or other software) as long as the heritage interface (red line) is provided by the MOSA
components. This flexibility is helpful in rolling a new architecture into an existing system.
Other added functionality, shown on the right side of Figure 12, is also possible in a MOSA system. In this case, the link
between A and B is replaced by software applications that look like A and B, but enable splitting out As interactions
into multiple B units, or adding a logging function to the network traffic.

Proc. of SPIE Vol. 9096 90960K-10

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 03/06/2015 Terms of Use: http://spiedl.org/terms

If designed correctly, MOSA systems can take any digital

communication and jump in the middle
Working within heritage system
Herii age Components

Improving capabilities or scope of system

Ungi nal:

Expanded, networked capability:

MOSA
MOSA
OSA boundary

Logger

MOSA Components
Figure 12. MOSA-Based System

4. CONCLUSION
The overarching objective for Flexible Weapons is to replace current inventory weapons that will not fully utilize the
increased capabilities of 6th generation platforms, with a single weapons kit made up of flexible, open architecture
components. Flexible Weapon will develop a common architecture to enable modular subsystems to achieve flexible
weapons capability while allowing technology refresh at the pace of technology discovery in an affordable and
sustainable design. The various combinations of weapons to address multiple missions must be 100% compatible with
6th generation delivery platforms (fighters, bombers, RPAs) and backwards compatible with 4th and 5th generation
platforms.

Proc. of SPIE Vol. 9096 90960K-11

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 03/06/2015 Terms of Use: http://spiedl.org/terms