Unit -1 Embedded systems

Block diagram of General purpose computer system

Control ALU
Input unit unit Output unit

Registers

Main Memory
(e.g RAM)

Secondary Memory
(e.g Hard Disk)

A computer is a combination of two components namely hardware and software. Hard disk, CPU,
memory (volatile also called RAM), CD-ROM and printers are a few examples of the hardware which are
considered as part of a computer. The software are applications for various purposes and operating systems
(Windows and Linux are two highly used operating systems) to manage the applications and computer
hardware efficiently. Desktops and laptops are examples of such digital computer systems.

Input devices, such as keyboard and mouse, allow a user to provide input to the computer system.
The users can see useful information on output devices such as monitors and printers. These input/output
(I/O) devices communicate with the CPU through different types of interfaces. The CPU carries out all
kinds of processing and comprises three basic building blocks, namely, arithmetic logic unit (ALU), control
unit and set of registers.

The main memories are random access memory (RAM) and read-only memory (ROM). These
memories are accessible directly from the processor. Low access-time and random access are their major
attributes. Hard disk is an example of the secondary memory and does not interact with the CPU directly.
Its information access-time is relatively higher than that of the primary memories

1 |Embedded systems
EMBEDDED SYSTEMS

A generic block diagram of an embedded system is shown in Figure

Microprocessor

Memory

Microcontroller

Peripheral Devices

Embedded systems

Electrical,
I/O devices (Display, keypad)mechanical, optical, etc. device

An embedded system is developed for a specific application to perform dedicated task(s). An

embedded system consists of a processor, memory and other interfaces and is the building blocks of a
microcontroller which is embedded inside a specific embedded system. The software is designed to handle
a range of different tasks and is normally programmed to ROM memory. The ROM in embedded systems is
employed to store programs permanently even when the power is turned off.

Advantages:

 small size,
 low cost, and
 low power requirements

Examples of Embedded system:

 cell-phones and other handheld devices,

 blood pressure monitoring equipment
 digital multi-meters
 temperature and humidity measuring devices

2 |Embedded systems
 From the system hardware perspective, an embedded system embeds a microcontroller inside it, which
in turn contains a microprocessor core. This is a generic hierarchical view

Embedded system

Microcontroller

Microprocessor

MICROCONTROLLERS

1. Applications in communication: Radio, telephone, cellular phones, answering machines,fax

Machines, wireless routers.
2. Consumer electronics: Washing machine, clocks and watches, games and toys, remote controls,
Audio/video electronics.
3. Automotive systems: Braking system, electronic ignition, locks, power windows and seats,
collision avoidance
4. Commercial usage: ATM machines, bar code readers, elevator controllers.
5. Medical treatments: Cancer treatments, dialysis machines, blood pressure measuring equipment,
(ECG), etc.
6. Industrial: Process automation, oil refineries, food processing plants, paper and board mills, etc.
7. Military use: Missile guidance systems, global positioning systems, surveillance systems.

Design Parameters of Embedded Systems

While designing an embedded system, the important factors to b considered are -

1. Power consumption
2. Speed of execution
3. System size and weight
4. Performance

Microcontrollers

A microcontroller combines a microprocessor, read only memory (ROM), random access memory (RAM),
and input/output (I/O) peripheral devices on a single chip. For this communication among subsystems,
different buses are used. A bus is a collection of wires over which digital signals propagate in the form of 0s

3 |Embedded systems
and 1s. Buses internal to the chip can be broadly categorized in three groups namely data, address, and
control.

The basic block diagram of a microcontroller is as shown below.

ADDR BUS DATA BUS

CONTROL BUS

Microprocessor

CODE MEMEORY

DATA MEMEORY

CONTROL & TIMING INTERFACES

DIGITAL I/OS ANALOG I/OS COMMUNICATION INTERFACES

Any memory location inside RAM, ROM and in registers is identified by unique numbers called
addresses.
Address signals flow over the address bus. The size of the address bus decides the unique addresses
generated by a microprocessor.
The data bus carries the data. Increasing the size of the data bus allows more number of bits to be
communicated between two subsystems.
Control signals such as read/write inform whether the CPU is interested in reading some
information or wants to write some information to the location whose address was generated over the
address bus.
Major component of a microcontroller is the general purpose peripheral device called port, which
provides a physical connection between the microcontroller and the outside world devices. Based on the
type of the device being connected with the microcontroller, ports can be configured as input or output. In

4 |Embedded systems
addition, some of the port registers are used to monitor the device status, while other registers are used to
configure the direction of the port.
Information exchange between a microcontroller and a device can take place in parallel or serially.
Broadly, I/O ports can be classified in two categories
• Parallel Port: When two communicating entities are connected with a group of lines, a number of
data bits can be exchanged simultaneously.
• Serial Port: In this case, there is only one line for the data to travel between two communicating
parties.
Difference between the Microcontrollers and Microcomputers
Microcomputers Microcontrollers
RAM, ROM, CPU and I/O interfaces are integrated RAM, ROM, CPU and I/O interfaces are embedded
in different chips on a motherboard inside a single chip
More speed of memory and other peripherals The speed of memory and other peripherals is less
Size of the memory is more Size of the memory is less

Memory: Information Storage Device

All Digital systems store information in binary digits called bits. These bits are used for representing
Operators, operands, and addresses. There are two states for a binary bit. A `1' represents the presence of a
voltage and a `0' represents the absence of voltage. This representation is called positive logic.
To store binary information, CMOS circuits can be built and a collection of a large number of such circuits
is called memory. Memory stores data sequentially which is as shown in Figure.

Memory Address Data

0x201A 3241 10110111
0x201A 3242 01010100
. .
. .
. .
. .
. .
.
0x201A 3244 1011 1101
0x201A 3243 1010 1101
0x201A 3242 1011 1100

Memories are made up of data storage locations, which are uniquely addressable and can be accessed by the
processor. Each storage location can hold 8 bits of information. Since each byte is uniquely addressable,
such memories are called byte addressable memories.

5 |Embedded systems
During a write operation, data is transferred from processor to memory. Whereas in Read operation, data is
transferred from memory to the processor
Based on read and write operations, memories are grouped in two categories namely
 Read-only memory (ROM) and
 Random access memory (RAM).
The ROM allows read operations only
In case of RAM both read as well as write operations can be performed by the microprocessor.
Read Only Memory
The memory that allows the processor to only read its contents is Read Only Memory (ROM).
Another characteristic of ROM is that they can store information permanently even when no power is
applied. The information in the ROM is therefore non volatile.
Instructions or a program code are stored in a ROM.
Example: The boot sequence with which an operating system is loaded.
Based on different methods of programming the ROM, ROM is classified into
 PROM
 EPROM
 EEPROM
 Flash memory
Programmable ROMs (PROMs):
These can be programmed only once. Any change to the contents will require the replacement of the
chip.
Erasable Programmable ROMs (EPROMs)
These can be reused by erasing and re-programming. Conventionally, information was erased and
written to with the help of ultraviolet (UV) light. The changes cannot, however, be made while the chip is
installed in the system. Another drawback of EPROM is that the whole chip needs to be completely erased.
Electrically Erasable PROMs (EEPROMs)
These are EPROMs but their contents can be erased and written to by applying electric signals to
the storage cells. Furthermore, erasure and writing can take place while the chip is installed in the circuit.
Flash memory
It is a type of EEPROM that allows read and write operations to be carried out in large multi-byte
blocks. In general, the erase cycles for non-volatile memories are slow. Flash memory allows erasing large
block sizes, which provides these memories a significant speed. One of the limitations of ash memories as
well as EEPROMs is their limited number of read and write cycles.

6 |Embedded systems
Random Access Memory
The memory which allows the processor to read from and write to its locations is known as Random
Access Memory (RAM).
One limitation of RAM is that information stored in it is lost as soon as the power applied to it is removed.
On the other hand, RAM is not limited by the number of read and write cycles and is more suitable for
storing data that is updated frequently. The read and write operations of RAMs are faster than those of
ROMs.
There are different types of RAMs namely-
 Dynamic RAM (DRAM)
 Static RAM (SRAM)

Dynamic RAM (DRAM) is the most commonly used type of RAM. Each memory cell of a DRAM,
can store one bit of information. It is made up of two transistors and a capacitor. The transistor acts as a
switch while the
Capacitor holds the charge. As the capacitor discharges, the voltage representing `1' needs to be refreshed.
This refresh operation is performed a number of times in one second and results in reducing the memory
operating speed.
Static RAM (SRAM) employs a flip-flop for storing a bit in a memory cell. A Flip-flop requires 4 to 6
transistors and does not require refreshing circuitry. SRAM is faster than DRAM because of the absence of
refresh cycles.
On a given chip SRAM yields less memory space as compared to DRAM. This is because more numbers
of transistors required by an SRAM cell. This also makes them more expensive. Because of their features,
SRAM is normally used for building cache memories while DRAM is employed for RAM chips. In
comparison to ROM, the access time of RAMs is less.
Differences between DRAM & SRAM

7 |Embedded systems
Aligned and Unaligned Memory Accesses
For a byte (8bit) addressable memory, each memory access of size byte is inherently aligned.
A 32-bit processor will have a 32-bit data bus to access the memory. It is possible to perform memory
read/write operations of half word (16-bit) and word (32-bit) size. If a memory access, of size halfword, is
performed from an even address, then this memory access is an aligned access. On the other hand, if an
odd address is used, then the resulting memory access is an unaligned access.
Similarly, for a 32-bit word memory access, word aligned means the data is stored on a word
boundary, i.e., the memory address accessed is divisible by 4. A word memory access from an address not
divisible by 4 is termed unaligned word access.
THE MICROPROCESSOR
A Microprocessor, also known as a central processing unit, has a capability of executing instructions at an
extremely high speed. Broadly, a microprocessor has five basic components namely –
 Arithmetic logic unit (ALU),
 Control unit,
 Bank of registers,
 Interconnection buses, and
 Timing unit.
The block diagram of a microprocessor is shown in Figure.

The functions of various blocks of a

Program counter
microprocessor is described below

Arithmetic Logic Unit (ALU)

Control unit Arithmetic logic unit

Arithmetic logic unit (ALU) is to perform
arithmetic operations such as addition,
subtraction and logical operations such as,
such as AND, OR operations.
Timing Unit
Register file

Control unit:
As the Microprocessor works with only binary digits, each operation should be encoded by a unique
binary combination, which is termed the operational code or the op-code. Various steps required for the
execution of an operation are performed using binary signals, which are known as control signals. These
control signals are generated by the control unit.

8 |Embedded systems
 The operations and operands are initially stored in the memory. They are brought inside the
microprocessor using buses and are stored temporarily in Microprocessor registers. Data transfers among
different processor registers as well as ALU are performed using the buses inside the microprocessor.
The function of a microprocessor is to execute a user program. The executable program is stored in
memory initially and is executed instruction by instruction. During program execution, the microprocessor
performs different operations.
The instructions from memory and data from memory or I/O ports is stored temporarily inside the
microprocessor using registers.
Working of a microprocessor
The working of the microprocessor is as shown in the following figure.

Memory
R=X+Y
Let us assume that the microprocessor
X=5
Y=7 executes R = X + Y addition instruction. Let's
assume that this instruction is compiled and is
Control Unit
present in the memory. This instruction will
consist of three fields. One of them will
contain the operational code representing
addition, while the other two fields will
Register Register 2 contain the memory addresses of the locations
1
where variables X and Y are stored. The PC
Accumulator register contains its memory address.

The steps that are carried out by the microprocessor are listed below.
 The microprocessor fetches the instruction from the address that is present in PC on the address bus
 The control unit decodes its opcode
 Processor executes R = X + Y by
- fetching the current value of X from the memory
- fetching the current value of Y from the memory
- instructing the ALU to add these two numbers
- writing the sum back to the memory address of R
These are the steps carried out by the processor to execute one instruction.

9 |Embedded systems
Difference between Microprocessor and Microcontroller:

Sl.
Microprocessor Microcontroller
No
1 CPU is stand alone, RAM,ROM ,I/O devices
CPU, RAM, ROM,I/O devices and TIMERS
and timers are separate and interfaced with
are all located on a single chip.
CPU
2 Designer can decide on the amount of RAM,
Fixed amount of RAM,ROM and I/O ports
ROM and I/O ports.
3 Applications in which cost, space and power
Expensive applications
are critical
4 Versatile and general purpose Not very versatile
5 Uses different program memory and data
Program memory and data memory are same
memory.
6 Large number of instructions with flexible Limited number of instructions with few
addressing modes addressing modes
7 Very few instructions which have bit handling Many instructions with bit handling
capability capability
8 System cost is more System cost is less
9 Clock frequency > 1 G Hz Clock frequency 10-20 M Hz

Microprocessor Architecture Classification

Microprocessor

Instruction set Architecture Memory interface based Architecture

Complexity based classification Instruction operand based classification Harvard Architecture Von Numan
Architecture

CISC MEMORY- MEMORY REGISTER – REGISTER

REGISTER - MEMORY
RISC

The architecture of the microprocessor can be classified as-

1. Instruction Set Architecture (ISA)

2. Memory interface based Architecture

10 |Embedded systems
Instruction Set Architecture (ISA)
The Instruction set Architecture is further subdivided based on
 Complexity of instructions and
 Instruction operands.
Complexity-Based ISA classification
Based on the ISA complexity the microprocessors are categorized into two groups:
1. Complex instruction set computer (CISC) and
2. Reduced instruction set computer (RISC).
The differences between RISC & CISC architecture is as shown below

RISC Architecture CISC Architecture

Complex Tool Development Software Tools Simplified Tool Development

Simplified Hardware Processor Complex Hardware

Complex instruction set computers (CISC):

A complex instruction set computer (CISC) is a computer architecture in which single
instructions can execute several low-level operations (such as a load from memory, an arithmetic
operation, and a memory store) or are capable of multi-step operations or addressing modes within
single instructions.

The key features of the CISC architecture are discussed below

 This architecture is suited where the processor speed faster than available memories
 A single complex instruction can perform many operations.
 Complex instructions require more clock cycles to complete
 Most of the instructions can access memory.
 A program running on a CISC architecture involves a small number of complex instruction.
 In CISC the complexity is embedded in the processor hardware, making the compilation tools
design simpler.
Example: Intel (x86) and Freescale 9S12.

11 |Embedded systems
Reduced Instruction Set Computers (RISC):
A Reduced Instruction Set Computer is a type of microprocessor architecture that utilizes a
small, highly-optimized set of instructions rather than the highly-specialized set of instructions.
The key features of the RISC architecture are discussed below
• This architecture is suited where the processor speed matches that of memories.
• Simplified instructions set.
• Each complex operation is broken into multiple simplified operations and dedicated instructions are
provided for these operations.
Example: load and store instructions
• As the instructions set are simple they are executed in a single processor clock cycle.
• Requires fewer memory addressing modes.
• Simple hardware architecture.
Example: MIPS, ARM, SPARC and PowerPC.
Difference between the RISC and CISC Processors

RISC CISC
It is a Reduced Instruction Set Computer. It is a Complex Instruction Set Computer.
It requires multiple register sets to store the It requires a single register set to store the
instruction. instruction.
RISC has simple decoding of instruction. CISC has complex decoding of instruction.
It uses a limited number of instructions It uses a large number of instructions
It uses LOAD and STORE in the register-to-register It uses LOAD and STORE instruction in the
a interaction of program. memory-to-memory interaction of a program.
RISC has more transistors on memory registers. CISC has transistors to store complex instructions.
The execution time of RISC is very short. The execution time of CISC is longer.
It has fixed format instruction. It has variable format instruction.
The program written for RISC architecture needs to Program written for CISC architecture tends to take
take more space in memory. less space in memory.
Example: ARM, Power Architecture, Alpha, AVR, Example: VAX, Motorola 68000 family and the
ARC and the SPARC. Intel x86 CPUs.

Instruction Operand-Based ISA Classification

The operands for an instruction can be specified either using memory or registers or combination of both.
The ISA classification based on how the operands are specified can be categorized in the following groups.
1. Memory-memory:
This type of ISA allows more than one operand to be specified in memory.
Example: VAX and PDP series.

12 |Embedded systems
 2. Register-memory:
These architectures allow one operand of an instruction to be specified in memory, while the other
operand is in CPU register. In this ISA the individual instructions execute faster. However, this may
require more number of instructions to complete the same task.
Example: x86 and Motorola 68k.
3. Register-register:
This ISA classification is also called load-store architecture. Memory is accessed using load and
store instructions. All instructions other than load and store instructions get their operands and store
their results to registers. The execution of these instructions is very fast. But this ISA requires the
most number of instructions to complete a given task. Since in many cases the immediate results are
not stored to the memory, and are temporarily placed in the registers, it might be used by the
subsequent instructions.
Example: ARM and MIPS.
Memory interface based Architecture classification
There are two types of memory interface architectures, namely,
 Von Neumann Architecture and
 Harvard Architecture
Von Neumann Architecture
The following figure shows the architecture of Von Neumann Architecture

Address Bus The von Neumann architecture uses a

Data Bus
common bus for both data and code
Control Bus memory. As a result either an instruction
Processor
can be fetched from memory or data can
be read/written to/from memory during
each memory access cycle. Instructions
Code Memory and data are stored in the same memory
subsystem and share a common bus to
the processor.
Data Memory
Example: Micrprocessors

13 |Embedded systems
Harvard Architecture
The following figure shows the architecture of Harvard Architecture
Address Bus

Data Bus
Control Bus

PROCESSOR

CODE MEMORY DATA MEMORY

The Harvard architecture, uses separate buses for accessing code and data memories. Hence
data and instructions can be accessed simultaneously. In addition, the next instruction may be fetched
from memory at the time when the previous instruction is about to finish its execution. However, main
memory access time is a major bottleneck in the overall performance of the system.
Example:. ARM-based microcontrollers.
Difference between Von Neumann Architecture and Harvard Architecture

14 |Embedded systems
Performance Comparison of Different Architectures
A microprocessor’s architecture can be compared against other architectures by running benchmark
programs. The following performance evaluation equation is used to quantify the execution speed of a
microprocessor.
N
Execution Time = Tc∑ Ci - - - - - - -- - - - -(1.1)
i=0

In (1.1), Tc is the cycle time, Ci represents the number of cycles required for i th instruction execution,
and N is the number of instructions in the test program. If each instruction requires the same number of
cycles for execution, then the expression in (1.1) is simplified to
Execution Time = Tc N C.- - - - - - - - - - - - - -- - - - - -- (1.2)
From the expression in (1.2) we observe that there are three different possible ways to speed up the
execution of a microprocessor and are listed below:
1. Use fewer instructions for a given program. In other words, it is possible to improve the
execution speed by efficient programming.
2. Reduce the number of cycles for the instructions. This is mainly dependent on the instruction set
architecture of the microprocessor.
3. Speed up the clock frequency of the microprocessor or equivalently reduce the cycle time. This
refers to the maximum clock frequency of the processor. The maximum clock frequency
depends on –
a) The physical limitation of processor clock frequency
b) The technical limitation associated with increasing the clock frequency

a) The physical limitation of processor clock frequency:

Let us assume a microprocessor is running at 3 GHz clock frequency. i.e., one clock cycle takes

1 1
ns. Now the electrical signals travel a physical distance of approximately 10 cm during ns.
3 3
Assuming that a circuit is working synchronously, it is necessary that a clock signal is available at
different parts of the circuit simultaneously. This can be ensured by requiring the propagation delay
much less than clock cycle time. Let us use a factor of 10, which will require the size of the circuit
to be about 1 cm. The CPU core die size of Intel Core2 Duo Processor is approximately 1 cm. Now
if we want to double the speed of this processor, the size of the processor core should become one-
half. If this size limitation is violated, then some parts of the processor circuit are operating in the
current clock cycle, while some other parts of the circuit are still not done with the previous clock
cycle. This type of distributed system is very hard to deal with in practical systems.

15 |Embedded systems
 b) The technical limitation associated with increasing the clock frequency.
As the CPU cores become smaller and smaller, the problem of heat dissipation starts arising. As
processors are made up of transistors and these transistors can be switched ON and OFF at the clock
frequency, each transistor dissipates power when switched from one state to another. Switching at fast
speed leads to more power dissipation. When the processor core size is reduced to increase the clock
speed, the corresponding amount of heat generated by these switching transistors couldn't be dissipated
easily due to the reduced chip size. The amount of speed gain and the corresponding increase in the cost
of associated cooling mechanism are some of the practical barriers in increasing the processor speeds
further. One natural solution to this limitation is to use multiple processors of moderate speed rather
than to have one processor of too high speed, leading to multi-core processor architecture.

Software System and Development Tools

The intelligence of a microcontroller-based embedded system lies in its software which is a collection of
instruction sequences, also called functions, that are stored in memory (either ROM or RAM). These
functions are implemented to perform certain tasks and are fundamental building blocks of a software
system. Based on the collective functionality of a set of functions, the software subsystem includes -
• Operating system,
• Device drivers, libraries and
• User applications.
Operating system:
Operating system includes the microprocessor, memory, Input/Output devices. The operating
system manages these devices using the hardware architecture dependent OS components. Also it allows
user applications to use these devices in a systematic way through OS. These interfaces provided by the OS
is as shown in Figure

User Application 1 User Application k

Operating System Call interface for Application

Operating System (OS) core System Libraries

Architecture Dependent OS Components Device Drivers

Hard ware

16 |Embedded systems
Device drives or drivers:

A device driver is a software sub-system, which allows the operating system to communicate with the
hardware devices. Drivers for some of the standard devices are sometimes integrated as an essential
component of the operating system. But for some hardware devices, the drivers are provided by the vendor.

Libraries:
The software library is a collection of function calls developed for specific job and is made available to
the user. By using a library, the software application developer reuses the specific functionality already
implemented by the library. Libraries allow sharing the code in a modular fashion, as well as easing the
distribution of the code.

Applications:
Application is a collection of one or more programs, which are designed to perform operations for
a specific application requirement. Application software cannot run (or execute) on its own. It is dependent
on the OS and libraries.

Software Development Tools

The process of converting user program to executable outputs involves two main steps:
(1) Compilation and
(2) Linking.

Compilation Process
The tool used in the compilation process is called either a compiler or an assembler depending on the type
of the user program source file.
If the user program is written in assembly language, then an assembler is used to convert source
code to an object code, which is also called machine code. In this process, an assembly instruction is
converted to its equivalent opcode or machine code. Assembly language programs are also called low level
programs.
If the user program is written in a high level language, then a compiler is used to convert it to the
machine code. Most of the compilers also provide the assembly equivalent code, in addition to machine
code. The process of converting a user program to machine code using either a compiler or an assembler is
as shown in Figure
17 |Embedded systems
 Source Code (High level)

Compiler

User assembly code Source Code (Low level)

Assembler
Assembler

Machine Code Machine Code

The selection of a compiler is not only dependent on the choice of the high level language, but also on the
hardware platform to be used for running the program.
For example, a program is written in C language and the user wants to run its corresponding executable on
an Intel microprocessor then we need to compile the C language program using Intel's C compiler.

Building an Executable Using a Linker

The job of a linker is to construct an executable by combining different object codes (or object files)
obtained after compiling the user program codes. In the process of linking, the linker has to decide the
locations (addresses) of different object codes and data segments.
In addition, some of the functions used in the program are not implemented by the programmer, rather they
are provided by a library. It is the job of a linker to get the desired modules from different libraries and
integrate them in the process of constructing the executable. The functioning of the linker is shown in
Figure
Source code Source code
(High Level) (High Level)

Compiler Compiler

Machine/ Machine/
Object code Object code

Configurations Library
Linker

18 |Embedded systems
Executable
MAP file
List file
The content of a library is nothing but a collection of different object codes and the library itself is
constructed using a linker. The linker, in addition to building the executable, can generate, many other files
containing useful information for the programmer. Similarly, the list file can also be generated by the tools,
which provides the disassembly of the code which is helpful to the programmer in debugging.
The MAP and list files optional outputs from the linker. Another important related concept is the
dynamic linking, where the libraries are not used at the time of building an executable, rather they are used
during the execution at runtime and are called dynamic link libraries or in short DLLs.

19 |Embedded systems