Global Journal of Advanced Engineering Technologies, Vol1, Issue4-2012

ISSN: 2277-6370

VLSI IMPLEMENTATION OF PHYSICAL LAYER

CODING USED IN SUPER SPEED USB USING
VERILOG
K.V.Suresh babu1 , Prof.A.S.Srinivasarao2 , D.Srinivasa Rao3
1
Student (M.Tech), Dept of ECE, AITAM, Tekkali, A.P
2
Professor &H.O.D., Dept of ECE, AITAM, Tekkali, A.P
3
Assistant. professor, Dept of ECE, AITAM, Tekkali, A.P

Abstract- This project implements the DC balanced 8B/10B

coding used in Super speed USB with employ a very fast FPGA
from Xilinx family is proposed. This technique can be used by
other high speed serial buses such as PCI Express, IEEE
1394b, Serial ATA, SAS, Fiber Channel, SSA, Gigabit
Ethernet, In fini Band, XAUI, Serial Rapid IO, DVI and
HDMI (Transition Minimized Differential Signalling)that use
the same coding. Using the look-up table and memory with
fast technique made this design efficient to be implemented.
Moreover, the proposed method has very low
complexity and fast to execute with minimum logic and also
easy to implement. The Scrambling and descrambling
modules are added in the above modules to support USB 30
physical layer transactions. The code is implemented in
VERILOG
keywords Usb, link layer, crc, ncverilog and cadence
I. INTRODUCTION
It is a specification to establish communication between
devices and a host controller (usually a personal computer),
developed and invented by Ajay Bhatt, while working for
Intel. USB has effectively replaced a variety of interfaces
such as serial and parallel ports. USB can connect computer
peripherals such as mice, keyboards, digital cameras,
printers, personal media players, flash drives, Network
Adapters, and external hard drives. For many of those
devices, USB has become the standard connection method.
A. Version history:
USB 0.7: Released in November 1994.
USB 0.8: Released in December 1994.
USB 0.9: Released in April 1995.
USB 0.99: Released in August 1995.
USB 1.0 Release Candidate: Released in November 1995.
USB 1.0: Released in January 1996.Specified data rates of
1.5Mbit/s (Low-Bandwidth) and 12Mbit/s (Full
Bandwidth). Does not allow for extension cables or passthrough monitors (due to timing and power limitations).
Few such devices actually made it to market.
USB 1.1: Released in September 1998.Fixed problems
identified in 1.0, mostly relating to hubs. Earliest revision
to be widely adopted.
USB 2.0: Released in April 2000.Added higher

maximum bandwidth of 480 Mbit/s [60 MB/s] (now called

"Hi-Speed").Micro-USB
Cables
and
Connectors
Specification 1.01: Released in April 2007. Motivation
Connection of the PC to the telephone: Ease-of-use:
Objective of the Specification.
II. INTRODUCTION TO USB 3.0
This paper is a brief review of the USB 3.0 implementation,
focusing on USB 2.0backward compatibility and on the
major features associated with the Super-Speed (SS) bus.
The goal is to provide the reader with a short and concise
description of USB 3.0, and enough detail to give a good
feel for the technology, protocols, and techniques. Due to
the limited scope of this paper, some terminology and
concepts are introduced but not fully developed.
a) Motivation for USB 3.0:
USB 3.0 enables more demanding applications
compared to USB 2.0 by addressing its limitations:
Bandwidth - 5.0 G b/sec Super Speed (SS) vs. 480
Mb/sec (High Speed) rate.
Power Conservation - link power states (U0 - U3)
and function power management.
Data Flow Control - poll once versus poll multiple times
Error Handling - End-to-end and port-to-port error
detection and retries versus only end-to-end retries with
USB 2.0.The additional bandwidth provided by USB SS
transactions can benefit applications like real-time audio
and video streaming that require higher bus bandwidth
at regular intervals. Mass storage applications can also
benefit from the SS bandwidth. For example, Table 2.1
lists approximate download times for the different
transmission rates.
b) USB 2.0 Host:
Hubs power downstream ports when directed by software
(USB 2.0) with Super Speed support disabled.
Device connects at high-speed.
Host system begins device enumeration at high-speed.
When the host is powered off, the hub does not
provide power to its downstream ports unless the hub
supports charging applications. When the host is
powered on and there is no Super Speed hardware
support, the following is the typical sequence of events:
Hub detects VBUS and connects as a high-speed device.

218

Global Journal of Advanced Engineering Technologies, Vol1, Issue4-2012

Host system begins hub enumeration at high-speed.
III. PHYSICAL LAYER
The physical layer defines the PHY portion of a port and
the physical connection between a downstream facing port (on
a host or hub) and the upstream facing port on a device. The
Super Speed physical connection is comprised of two
differential data pairs, one transmit path and one receive path.
The nominal signalling data rate is 5 Gbps.

Express protocol and signalling. This includes features such as;

data serialization and de-serialization, 8b/10b encoding, analog
buffers, elastic buffers and receiver detection
Some key features of the PCI Express PHY are:
Standard PHY interface enables multiple IP sources for
PCI Express Logical Layer and provides a target
interface for PCI Express PHY vendors.
Supports 2.5GT/s only or 2.5GT/s and 5.0 GT/s serial data
transmission rate.
Utilizes 8-bit, 16-bit or 32 -bit parallel interface
to transmit and receive PCI Express data.
Allows integration of high speed components into a
single functional block as seen by the End point device
designer.
Data and clock recovery from serial stream on the PCI
Express bus.
Holding registers to stage transmit and receive data.
Supports direct disparity control for use in transmitting
compliance pattern(s).
c)

Figure 1: Super Speed Physical Layer

The electrical aspects of each path are characterized

as a transmitter, channel, and receiver; these
collectively represent a unidirectional differential
link. Each differential link is AC-coupled with
capacitors located on the transmitter side of the
differential link. The channel includes the electrical
characteristics of the cables and connectors. At an
electrical level, each differential link is initialized by
enabling its receiver termination.
a) Physical Layer Overview:
The physical layer defines the signalling technology for the
Super Speed bus.
This chapter defines the electrical
requirements of the Super Speed physical layer. This section
defines the electrical-layer parameters required for
interoperability between Super Speed components. The PHY
Interface for the PCI Express and USB Super Speed
Architectures (PIPE) is intended to enable the development of
functionally equivalent PCI Express and USB Super Speed
PHY's. Such PHY's can be delivered as discrete IC's or as
macro cells for inclusion in ASIC designs. The PIPE
specification is defined to allow various approaches to be used.
Where possible the PIPE specification references the PCI
Express base specification or USB 3.0Specification
rather than repeating its content. In case of conflicts, the
PCI-Express Base Specification and USB 3.0 Specification
shall supersede the PIPE spec. This spec provides some
information about how the MAC could use the PIPE interface
for various LTSSM states and Link states.
b) PCI Express PHY Layer:
The PCI Express PHY Layer handles the low level PCI

ISSN: 2277-6370

Physical Layer Transmission and Reception:

Transmission:
Scrambling- Scrambling reduces EMI problems
associated with repeated patterns in the data being
sent across an SS link. The scrambler output is simply
XORed with each byte of data to eliminate the repeated
patterns.
8/10b Encoding every byte that traverses the link is first
Converted into a10-bit value called a symbol (this is
a common encoding scheme in high-speed serial designs).
Parallel/Serial Conversion Bytes are converted to
bit stream LFPS Low Frequency Periodic Signaling
is typically used in situations where the link is in an
electrical idle state.
Differential Transmission Packets are clocked onto the
link at a 5.0 Gb/srate.
Reception:
Differential Reception the scrambled and encoded data
is received and forwarded to the recovery blocks.
Clock and Data Recovery the clock is extracted from
the bit stream and data is clocked into the serial/parallel
converter.
Serial/Parallel Conversion data is clocked into the
Converter and 10-bit symbols are clocked into the
elastic buffer.
Elastic Buffer The elastic buffer must absorb the worstcase clock variation between the transmitted
clock frequency (recovered) and the local receive
clock. The maximum variance is +300 to -300ppm. The
buffer must also accommodate variations resulting
from the Spread Spectrum clocking. Compensation is
achieved via SKP ordered sets that are periodically
inserted into the bit stream.
8/10b Decoding 10-bit symbols are converted back to

219

Global Journal of Advanced Engineering Technologies, Vol1, Issue4-2012

bytes. Un Scrambling the same scrambling output is
XORed with the scrambled data a second time to
recover the original data Rx Clk.
d) VLSI Implementation of physical coding layer
USB 3.0 ARCHITECTURE

Test bench:
The term test bench usually refers to simulation code used
to create a predetermined input sequence to a design, then
optionally to observe the response.
Today, in the era of multi-million gate ASICs and FPGAs,
reusable intellectual property (IP), and system-on-a-chip
(SoC) designs, verification consumes about 70% of the
design effort. Design teams, properly staffed to address the
verification challenge, include engineers dedicated to
verification. The number of verification engineers can be
up to twice the number of RTL designers. Higher abstraction
levels are usually accompanied by a reduction in control and
therefore must be chosen wisely. These higher abstraction
levels often require additional training to understand the
abstraction
mechanism and how the desired effect can be
produced. If a verification process uses higher abstraction
levels by working at the transaction- or bus-cycle levels (or
even higher ones), it will be easier to create large amount of
stimulus. But it may be difficult to produce a specific sequence
of low-level zeroes and ones.
ENCODER:
One of the major goals of 8b/10b encoding is to embed
a clock into the serial bit stream before transmission
across the link. This eliminates the need for a high
frequency 5.0 GHz clock signal on the link that could

ISSN: 2277-6370

generate significant EMI. Every byte to be sent is

converted to a 10-bit value, called a symbol a look-up table
associated with the encoder.
Data bytes consisting of every byte send across the
link except ordered Sets. The data lookup table must
support the 256 possible input values.
Control bytes used in ordered sets. The 8b/10b
encoding scheme is also designed to prevent DC
wander, meaning the possibility that too many bits of
one polarity could interfere with the ability of the
receiver to properly see them.
DECODER:
The 8b/10b Decoder uses two lookup tables (the D and K
tables) to convert the 10-bit symbol stream back into
bytes. Each symbol value is submitted to both lookup
tables but only one of the tables will find a match for the
symbol. The state of the D/K# signal indicates that the
received symbol is a:
Data (D) Symbol a match for the received symbol is
located in the D table. D/K# is driven High.
Control (K) Symbol a match for the received symbol
is located in the K table. D/K# is driven Low.
SCRAMBLER:
Scrambling reduces repeated patterns in the bit stream
and lowers EMI by preventing the concentration of
emitted energy at only a few frequencies. Scrambling
works by generating a pseudo-random data pattern that is
XORed with the outgoing bit stream. . The algorithm used
for scrambling data is expressed as a polynomial
implemented as a linear feedback shift register (LFSR).
DESCRAMBLER:
Disabling scrambling is intended to help simplify test
and debug equipment. Control of the exact data patterns is
useful in a test and debug environment. Since scrambling
is reset at the physical layer, there is no reasonable way to
reliably control the state of the data transitions through
software. The Disable Scrambling bit is provided in the
training sequence for this purpose. The mechanism(s)
and/or interface(s) used to notify the physical layer to
disable scrambling is component implementation specific
and beyond the scope of this specification.
IV.RESULTS
TOOLS USED FOR EXPERIMENTAL ANALYSIS: SIMULATORMODELSIM 6.3F, SYNTHESIS- ISE 9.2I, DEBUGGING TOOLCHIP SCOPE 9.2I, BOARD- SPARTAN 3E.

Figure 2: Clock and Reset generator signals

220

Global Journal of Advanced Engineering Technologies, Vol1, Issue4-2012

ISSN: 2277-6370

An synchronous clock with a total period of 20 ns or

50MHz was generated in the top level block for
synchronous operation, Asynchronous reset is generated for
negative edge Reset.

Figure 5: Port and internal logic signals of Encoder module

Generated 8 bit data from scrambler is given as input

encoder module and is observed as din signal, kin signal,
the result will be observed in dout which a 10 bit output from
the above figure.
Figure 3: Test bench generator signals

Using $random different values for input is generated and it

is sent via top module to Scrambler module in the
Architecture as shown in fig 3 .The results for randomly
generated input is shown above.

Figure 6: Port and internal logic signals of Decode module

Figure 4: Port and internal logic signals of Scrambler module

221

Global Journal of Advanced Engineering Technologies, Vol1, Issue4-2012

Fig. 7 Port and internal logic signals of Descrambler module

VI CONCLUSIONS
The Architecture of FPGA implementation of 8b/10b
coding used in super-speed USB is proposed and designed for
digital hardware implementation. All individual modules
have been designed individually and verified functionally
using random test bench using ModelSim 6.3f. It is observed
that the simulation results for the 8b 10b encoder, 8b 10b
decoder, scrambler and descrambler generated were
satisfactory and also the interconnections among all the
modules are perfect. A priority encoder method is used for 8b
10b encoder and decoder. The symbol errors for both D and
K symbols are verified in both encoder and decoder. The
entire above module are coded in VERILOG hardware
description language.
This work can extended by connecting this total module
in between Link layer and Physical analog layer of USB 3.0
architecture and transferring USB3.0 packets rigorously
from link layer to physical layer. Also the work can be
extended to do the FPGA implementation by using SPRTAN
3E or Vertex V XILINX FPGAs. The above project is best
suits as an IP (Intelectual Property) core of Physical Coding
Layer in USB 3.0 specification.
VII FUTURE SCOPE
The Design Physical coding layer can be rigorously
tested if soft cores of link layer, physical analog layer are
available. This can be an extended work for the present
project.
Complete USB 3.0 project can be done if one has
finished soft cores of link, protocol, application layers in
addition to our project. If analog PHYs are available in the
market hardware level validation can be done to our project.

ISSN: 2277-6370
REFERENCES

[1] A.X. Widmer and P.A. Franaszek, A dc-balanced,

partitioned block, 8B/10B transmission code, IBM Journal
of Research and Development, vol.27, no.5, pp. 440-451,
Sep 1983.
[2] Universal Serial Bus 3.0 Specification Revision0.9, July30,
2008.
[3] Ravi Budruk, Don Anderson & Tom Sanely, 2004.
PCI Express System Architecture - Mindshare Inc., pp 419434.
[4] Thatcher, Jonathan (1996-04-01). "Thoughts on Gigabit
Ethernet Physical". IBM. Retrieved on 2008-08-17.
[5] Jenming Wu & Yu-Ho Hsu ,8B/10B Codec for Efficient
PAPR Reduction in OFDM Communication Systems.
[6] J.Begeron, what is verification? in writing Test
benches: functional verification of HDL models.
[7] International technology roadmap for
semiconductors(ITRS)
[8] http://www.doulos.com/knowhow/sysverilog/ovm/tutorial
[9] http:// www.ovm.world.org/white_papers.php
[10] A.Shankara Rao and D.Rama Devi, FPGA
implementation of 8b/10b coding used in super-speed USB
in an international conference on CEMC-2011, ILAHIA
College Of Engineering, Kerala. (Communicated).

222