a novel approach of caching direct mapping using cubic approach

21
A Novel Approach Of Caching Direct Mapping Using Cubic Approach Guidance Professor Chandrasegar .T Site,vit university Vellore Team Member Kamlesh Asati 11MCA0133 Aakansha Soni 11MCA0080 Harikesh Dwivedi 11MCA0083 04SETMCAO056

Upload: kartik-asati

Post on 08-Jul-2015

123 views

Category:

Education


2 download

DESCRIPTION

Hi.., I had work out on this research paper in duration my master degree(MCA) and Successfully present at 6th International Conference under "Science Engineering Technology (SET) " in may-2013 in VIT University.

TRANSCRIPT

A Novel Approach Of Caching

Direct Mapping Using Cubic

Approach

Guidance

Professor Chandrasegar .TSite,vit university Vellore

Team Member

Kamlesh Asati 11MCA0133

Aakansha Soni 11MCA0080

Harikesh Dwivedi 11MCA008304SETMCAO056

Aim

A novel approach of caching direct mapping

using cubic approach appreciate to improve

system efficiency without losses the data.

It means that decrease the data access time

with using sufficient memory but that’s not

enough while we are implement these things

with minimum time complexity and high

level security.

04SETMCAO056

Objective Scope

Increase the system performs.

Decrease the process execution time and find the method in which time complexity in this approach is minimum.

Try to produce accurate result without data losses.

04SETMCAO056

Abstract In the last few decades there is lot of research

works carried out for the improvement of c p u.

Caching address mapping technique plays avital role for the system improvement.

Based on the number of hits occurred on thecaching leads to the effective utilization ofprocessor.

in general when the caching memory sizeincreases then the number of page faults/misses is going to get decreases between theprocessor and the system memory.

In this concept the proposed work is to deal withcaching direct mapping and to secure the data ina random and cubic based combinatorialapproach.

04SETMCAO056

Introduction

Cache Memory Mapping

The memory system has to quickly determine if a given address is in the cache.

Caching address mapping techniques plays a vital role for the system improvement. Based on the no of hits occurred on the caching leads to the effective utilization of processor.

There are three popular methods

1) Fully associative mapping

2) Set associative mapping

3) Direct mapping

04SETMCAO056

Direct Mapping

This is the most popular technique for

cache memory mapping.

If each block from main memory has only

one place it can appear in the cache, the

cache is said to be direct mapped.

Since a data block from ram can only be on

space line in the cache , it must always

replace the one block that was already

there.04SETMCAO056

Address Splits into two part address

Least significant w bits identify unique word in block.

Most significant s bits specify one memory block.

s bits divided into two parts tag line

s

a) Cache line address field r bits.

b) Tag field of s-r most significant bit.

S w

S-r r

04SETMCAO056

Cubic Approach

In cubic approach for Mapping we map thedata in random and cubic base approach.

For this we use the formula of cubic equation.

F(x)= ax3+bx2+cx+d

04SETMCAO056

Existing System In direct mapping, mapping is done with the linear approach.

For this we use hashing concept, If data is consist of n integersand we have k number of cells then the address where in datastored or retrieved is the ordinary hash function is Hash = n%k

H (k, i) = (h (k) + ci) mod m

Think how linear mapping works then the answer is if the firstlocation is not free, then it will go to the next location andcheck if the location is free or not, and so on until it findsempty blocks or can’t find at all. For retrieving the data, hashfunction and rehash function is used, retrieving is done byusing the hash function to find the key and check if the datawould coincide to the data needed. If not then rehash wouldbe needed. Until such time the correct location is found or anempty space is encountered which shows the data does notexists. In linear approach the searching takes so much timebecause in linear mapping it searches in all blocks of memorytherefore the speed is slow.

04SETMCAO056

Literature Review According to Zhenghong Wang and Ruby B. Lee ,

Caches ideally should have low miss rates and shortaccess times, and should be power efficient to systemat the same time.

Sincerely finding on efficient hits based oninformation losses in caches have also moves thesecurity issue for it. finally this concept shows thatthe cache architecture has low Miss Ratescomparable to a highly associative cache and shortaccess times and power efficiency close to that of adirect-mapped cache.

Additional benefits that the include cachearchitecture can bring, like fault method and hot-spotrevolution techniques.

04SETMCAO056

According to Mark D. Hill , cache is used toimprove the system cost, which have largememory but using the cache memory weincrease the access time of the system.

Cache is small in size so it holds some data at atime .cache memory hold that data which arefrequently used by the processor. For storingcache retained the recent referencedinformation.

Caches are successful due to temporal andspatial locality. Temporal locality means futurereferences are likely to be made to the samelocations as recent references, while spatiallocality suggests that future references are alsolikely to be made to locations near recentreferences.

Caches take advantage of temporal locality byretaining recently referenced Information.

04SETMCAO056

Methodology

Now we are trying to use our cubic approach tomap the memory. The family of cubic equation is

f(x) =ax3+bx2+cx+d. We use the function

h(x) = (ax 3+bx 2+cx+d) mod y

For finding the key element in cache memory .inthis formula we take the value of x between 0 to 15and map the memory into cache between 0 to 15.

For finding the solution we gives the value of xfrom 0 to 15 and change the values of a, b, c, d andcalculate the function value y= ax3+bx2+cx+d andfor the value of key element we calculate the valuey mod m here m is 16.

04SETMCAO056

In this table weobserve this cubic

formulaF(x)=(ax3+bx2+cx+)mody gives thevalue which isrepeated in thecolumn (ymod16),itmeans for thatvalue of a,b,c,dit maps somememory addresswhich is same.sothis is not acorrect mapping.

X a b c d y Mod y

0 2 3 3 1 1 1

1 2 3 3 1 9 9

2 2 3 3 1 35 3

3 2 3 3 1 91 11

4 2 3 3 1 189 13

5 2 3 3 1 341 5

6 2 3 3 1 559 15

7 2 3 3 1 855 7

8 2 3 3 1 1241 9

9 2 3 3 1 1729 1

10 2 3 3 1 2331 11

11 2 3 3 1 3069 3

12 2 3 3 1 3925 5

13 2 3 3 1 4941 13

14 2 3 3 1 6119 7

15 2 3 3 1 7471 1504SETMCA56

State diagram

Start

0,2,4,6,8 0,2,4,6,8 1,3,5,7,9

1,3,5,7,9

1,3,5,7,9 0,2,4,6,8

0 to 9

a

Final/hit

b

miss

c d

04SETMCAO056

We notice that the function

F(x)= (ax3+bx2+cx+d)mod y

give unique value

when we choose the value of a b & c such

that a & b must be even and c must be odd

but d consist any value between 0 to 9.

above given state diagram shows how the

function will work.

04SETMCAO056

Result

Using this approach we decrease the time

complexity of the direct mapping which is

good and complexity is o(n) , its also help the

decrease the process execution time.

input(0-15) output(0-15)

The proposed work is to deal with caching and

to secure the data in a random and cubic base

approach.04SETMCAO056

Cubic remapping system

f(x) =ax3+bx2+cx+d

So we are using the cubic approach which

maps the function more speedily than linear.

In linear approach power consumption is also

high using this approach we can also reduce

the power consumption.

Using cubic approach we get the time

complexity is o(n) .

04SETMCAO056

Consider this snapshot as sample output ...

04SETMCAO056

ConclusionA cache is to improve system cost performance by

providing the capacity of the large, slow memory

with an access time close to that of the small, fast

cache.

This simple approach implements a form of by

passing the different address value.

We develop a cubic approach to evaluate the

performance of the proposed direct mapped caching

algorithm our method enhances the reuse behavior of

the temporal data while improving the reuse of the no

temporal data by cubic associativity.

From our experimental results, we see that our

method is both fast and accurate.04SETMCAO056

Future Enhancement

Future reaches issue can give better solution

for this system , provided it overcomes the

problems like replacement strategy that can

better adopt to user access pattern and improve

the cache space utilization , security issue.

04SETMCAO056

THANK YOU