Context & Overview Previous works ◮ Optimized single band CSMA/CA ◮ Contention window [Xu2012] ◮ CSMA/CA - ECA (Enhanced Collision Avoidance) [Barcelo2009] ◮ Multiple channels based on CSMA/CA ◮ Multiplexing users through different channels [Chong2009] [Kwon2010] PhD Defense Baher Mawlawi November 26, 2015 13 - 63
Context & Overview Previous works ◮ Optimized single band CSMA/CA ◮ Contention window [Xu2012] ◮ CSMA/CA - ECA (Enhanced Collision Avoidance) [Barcelo2009] ◮ Multiple channels based on CSMA/CA ◮ Multiplexing users through different channels [Chong2009] [Kwon2010] ◮ Separating physically the control and the data planes [Basagni2009] PhD Defense Baher Mawlawi November 26, 2015 13 - 63
Context & Overview Previous works ◮ Optimized single band CSMA/CA ◮ Contention window [Xu2012] ◮ CSMA/CA - ECA (Enhanced Collision Avoidance) [Barcelo2009] ◮ Multiple channels based on CSMA/CA ◮ Multiplexing users through different channels [Chong2009] [Kwon2010] ◮ Separating physically the control and the data planes [Basagni2009] ◮ crowded : CSMA/CA still runs on a common channel and suffers from collisions between control messages. PhD Defense Baher Mawlawi November 26, 2015 13 - 63
Context & Overview Previous works ◮ Optimized single band CSMA/CA ◮ Contention window [Xu2012] ◮ CSMA/CA - ECA (Enhanced Collision Avoidance) [Barcelo2009] ◮ Multiple channels based on CSMA/CA ◮ Multiplexing users through different channels [Chong2009] [Kwon2010] ◮ Separating physically the control and the data planes [Basagni2009] ◮ crowded : CSMA/CA still runs on a common channel and suffers from collisions between control messages. ◮ low traffic : high rates users are penalized. PhD Defense Baher Mawlawi November 26, 2015 13 - 63
Outline 1. Context & Overview 2. M - CSMA/CA - RTS/CTS 3. Scheduled M - CSMA/CA - RTS/CTS 4. Joint PHY-MAC analysis 5. Conclusions & Perspectives
M - CSMA/CA - RTS/CTS PhD Defense Baher Mawlawi November 26, 2015 15 - 63
M - CSMA/CA - RTS/CTS PhD Defense Baher Mawlawi November 26, 2015 15 - 63
M - CSMA/CA - RTS/CTS PhD Defense Baher Mawlawi November 26, 2015 15 - 63
M - CSMA/CA - RTS/CTS PhD Defense Baher Mawlawi November 26, 2015 15 - 63
M - CSMA/CA - RTS/CTS Analytical approach ◮ Development of analytical model based on Markov chains under some assumptions [Bianchi98] : PhD Defense Baher Mawlawi November 26, 2015 16 - 63
M - CSMA/CA - RTS/CTS Analytical approach ◮ Development of analytical model based on Markov chains under some assumptions [Bianchi98] : ◮ No capture effect. PhD Defense Baher Mawlawi November 26, 2015 16 - 63
M - CSMA/CA - RTS/CTS Analytical approach ◮ Development of analytical model based on Markov chains under some assumptions [Bianchi98] : ◮ No capture effect. ◮ Failed transmissions only occur as a consequence of collision (perfect physical channel). PhD Defense Baher Mawlawi November 26, 2015 16 - 63
M - CSMA/CA - RTS/CTS Analytical approach ◮ Development of analytical model based on Markov chains under some assumptions [Bianchi98] : ◮ No capture effect. ◮ Failed transmissions only occur as a consequence of collision (perfect physical channel). ◮ All nodes are saturated, always having packets to send. PhD Defense Baher Mawlawi November 26, 2015 16 - 63
M - CSMA/CA - RTS/CTS Analytical approach ◮ Development of analytical model based on Markov chains under some assumptions [Bianchi98] : ◮ No capture effect. ◮ Failed transmissions only occur as a consequence of collision (perfect physical channel). ◮ All nodes are saturated, always having packets to send. ◮ For any given node, the probability of collision, p i , is constant and independent of the node’s collision history of the node and all other nodes. PhD Defense Baher Mawlawi November 26, 2015 16 - 63
M - CSMA/CA - RTS/CTS Analytical approach ◮ Development of analytical model based on Markov chains under some assumptions [Bianchi98] : ◮ No capture effect. ◮ Failed transmissions only occur as a consequence of collision (perfect physical channel). ◮ All nodes are saturated, always having packets to send. ◮ For any given node, the probability of collision, p i , is constant and independent of the node’s collision history of the node and all other nodes. ◮ The probability of collision does not depend on the backoff stage at which the transmission is made. PhD Defense Baher Mawlawi November 26, 2015 16 - 63
M - CSMA/CA - RTS/CTS Analytical approach ◮ Development of analytical model based on Markov chains under some assumptions [Bianchi98] : ◮ No capture effect. ◮ Failed transmissions only occur as a consequence of collision (perfect physical channel). ◮ All nodes are saturated, always having packets to send. ◮ For any given node, the probability of collision, p i , is constant and independent of the node’s collision history of the node and all other nodes. ◮ The probability of collision does not depend on the backoff stage at which the transmission is made. ◮ All users have same bitrates and same amount of time to transmit. PhD Defense Baher Mawlawi November 26, 2015 16 - 63
M - CSMA/CA - RTS/CTS Analytical approach ◮ Development of analytical model based on Markov chains under some assumptions [Bianchi98] : ◮ No capture effect. ◮ Failed transmissions only occur as a consequence of collision (perfect physical channel). ◮ All nodes are saturated, always having packets to send. ◮ For any given node, the probability of collision, p i , is constant and independent of the node’s collision history of the node and all other nodes. ◮ The probability of collision does not depend on the backoff stage at which the transmission is made. ◮ All users have same bitrates and same amount of time to transmit. ◮ All users are divided into sets and each set of users is assigned to a predefined sub-channel. PhD Defense Baher Mawlawi November 26, 2015 16 - 63
M - CSMA/CA - RTS/CTS Analytical approach 1-p i 1/CW i,0 1 1 1 1 (0,0) (0,1) (0,2) (0,CW i,0 -2) (0,CW i,0 -1) p i /CW i,1 (i-1,0) 2 p i /CW i,j π i = � m i − 1 1 + W min i + p i W min i k =0 (2 p i ) k 1 1 1 1 (i,0) (i,1) (i,2) (i,CW i,j -2) (i,CW i,j -1) p i /CW i,j+1 (m i -1,0) p i /CW i,m i 1 1 1 1 (m i ,0) (m i ,1) (m i ,2) (m i ,CW i,m i -2) (m i ,CW i,m i -1) p i 1/CW i,m i PhD Defense Baher Mawlawi November 26, 2015 17 - 63
M - CSMA/CA - RTS/CTS Analytical derivations Bianchi model M - CSMA/CA - RTS/CTS model n � P n (1 − π i ) Ni tr = 1 − P tr = 1 − (1 − π ) N i =1 Nπ (1 − π ) N − 1 1 − � n i =1 (1 − N i π i (1 − π i ) Ni − 1 ) P n P s = s = 1 − � n 1 − (1 − π ) N i =1 (1 − π i ) Ni E [ Payload information transmitted in a slot time ] E [ Payload information transmitted in a slot time ] S n = S = E [ Duration of slot time ] E [ Duration of slot time ] P s P tr L P n s × P n tr L = = P s P tr T s + P tr (1 − P s ) T c + (1 − P tr ) T id P n s × P n tr T n s + P n tr × (1 − P n s ) T n c + (1 − P n tr ) T id PhD Defense Baher Mawlawi November 26, 2015 18 - 63
M - CSMA/CA - RTS/CTS Model validation PhD Defense Baher Mawlawi November 26, 2015 19 - 63
M - CSMA/CA - RTS/CTS Model validation ◮ For any given node, the probability of collision is constant and independent of the node’s collision history of the node and all other nodes. PhD Defense Baher Mawlawi November 26, 2015 19 - 63
M - CSMA/CA - RTS/CTS Model validation ◮ For any given node, the probability of collision is constant and independent of the node’s collision history of the node and all other nodes. ◮ The probability of collision does not depend on the backoff stage at which the transmission is made. PhD Defense Baher Mawlawi November 26, 2015 19 - 63
M - CSMA/CA - RTS/CTS Performance evaluation ◮ Home-made event-driven matlab simulator models the protocol behavior. Packet payload 8184 bits MAC header 272 bits PHY header 128 bits ACK length 112 bits + PHY header RTS length 160 bits + PHY header CTS length 112 bits + PHY header Channel Bit Rate 72.2 Mbit/s Propagation Delay 1 µ s SIFS 10 µ s Slot Time 9 µ s DIFS 28 µ s Table : PHY layer parameters for 802.11n 20Mhz PhD Defense Baher Mawlawi November 26, 2015 20 - 63
M - CSMA/CA - RTS/CTS Performance evaluation ◮ Home-made event-driven matlab simulator models the protocol behavior. ◮ Saturation conditions, no PHY,... Packet payload 8184 bits MAC header 272 bits PHY header 128 bits ACK length 112 bits + PHY header RTS length 160 bits + PHY header CTS length 112 bits + PHY header Channel Bit Rate 72.2 Mbit/s Propagation Delay 1 µ s SIFS 10 µ s Slot Time 9 µ s DIFS 28 µ s Table : PHY layer parameters for 802.11n 20Mhz PhD Defense Baher Mawlawi November 26, 2015 20 - 63
M - CSMA/CA - RTS/CTS Performance evaluation ◮ M - CSMA/CA - RTS/CTS vs. Single band CSMA/CA - RTS/CTS PhD Defense Baher Mawlawi November 26, 2015 21 - 63
M - CSMA/CA - RTS/CTS Performance evaluation ◮ M - CSMA/CA - RTS/CTS vs. Single band CSMA/CA - RTS/CTS ◮ 3 RTS bands and 100 nodes, we can achieve : PhD Defense Baher Mawlawi November 26, 2015 21 - 63
M - CSMA/CA - RTS/CTS Performance evaluation ◮ M - CSMA/CA - RTS/CTS vs. Single band CSMA/CA - RTS/CTS ◮ 3 RTS bands and 100 nodes, we can achieve : ◮ Collision probability gain = 70 % PhD Defense Baher Mawlawi November 26, 2015 21 - 63
M - CSMA/CA - RTS/CTS Performance evaluation ◮ M - CSMA/CA - RTS/CTS vs. Single band CSMA/CA - RTS/CTS ◮ 3 RTS bands and 100 nodes, we can achieve : ◮ Collision probability gain = 70 % ◮ Saturation throughput gain = 30 % PhD Defense Baher Mawlawi November 26, 2015 21 - 63
M - CSMA/CA - RTS/CTS Performance evaluation ◮ M - CSMA/CA - RTS/CTS vs. Single band CSMA/CA - RTS/CTS ◮ 3 RTS bands and 100 nodes, we can achieve : ◮ Collision probability gain = 70 % ◮ Saturation throughput gain = 30 % ◮ Transmission delay gain = 40 % PhD Defense Baher Mawlawi November 26, 2015 21 - 63
M - CSMA/CA - RTS/CTS Performance evaluation ◮ M - CSMA/CA - RTS/CTS vs. Single band CSMA/CA - RTS/CTS ◮ 3 RTS bands and 100 nodes, we can achieve : ◮ Collision probability gain = 70 % ◮ Saturation throughput gain = 30 % ◮ Transmission delay gain = 40 % ◮ Packet drop probability is divided by ≈ 3 PhD Defense Baher Mawlawi November 26, 2015 21 - 63
M - CSMA/CA - RTS/CTS Performance comparision 100 90 80 PHY Upper bound MAC Upper bound Normalized throughput (%) Multiband (5 bands) 70 Single band 60 50 40 30 20 10 0 0 10 20 30 40 50 60 70 80 90 100 Number of Mobile Stations PhD Defense Baher Mawlawi November 26, 2015 22 - 63
M - CSMA/CA - RTS/CTS Discussion PhD Defense Baher Mawlawi November 26, 2015 23 - 63
Outline 1. Context & Overview 2. M - CSMA/CA - RTS/CTS 3. Scheduled M - CSMA/CA - RTS/CTS 4. Joint PHY-MAC analysis 5. Conclusions & Perspectives
Scheduled M - CSMA/CA - RTS/CTS System Model PhD Defense Baher Mawlawi November 26, 2015 25 - 63
Scheduled M - CSMA/CA - RTS/CTS System Model PhD Defense Baher Mawlawi November 26, 2015 25 - 63
Scheduled M - CSMA/CA - RTS/CTS System Model PhD Defense Baher Mawlawi November 26, 2015 25 - 63
Scheduled M - CSMA/CA - RTS/CTS System Model PhD Defense Baher Mawlawi November 26, 2015 25 - 63
Scheduled M - CSMA/CA - RTS/CTS System Model PhD Defense Baher Mawlawi November 26, 2015 25 - 63
Scheduled M - CSMA/CA - RTS/CTS System Model PhD Defense Baher Mawlawi November 26, 2015 25 - 63
Scheduled M - CSMA/CA - RTS/CTS Performance comparision 100 PHY Upper bound 90 MAC Upper bound Multiband (5 bands) 80 #Scheduler=2 (5 bands) Normalized throughput (%) #Scheduler=3 (5 bands) 70 Single band 60 50 40 30 20 10 0 0 10 20 30 40 50 60 70 80 90 100 Number of Mobile Stations PhD Defense Baher Mawlawi November 26, 2015 26 - 63
Scheduled M - CSMA/CA - RTS/CTS Synthesis PhD Defense Baher Mawlawi November 26, 2015 27 - 63
Scheduled M - CSMA/CA - RTS/CTS Synthesis for dense networks PhD Defense Baher Mawlawi November 26, 2015 28 - 63
Scheduled M - CSMA/CA - RTS/CTS Synthesis for unloaded networks PhD Defense Baher Mawlawi November 26, 2015 29 - 63
Outline 1. Context & Overview 2. M - CSMA/CA - RTS/CTS 3. Scheduled M - CSMA/CA - RTS/CTS 4. Joint PHY-MAC analysis 5. Conclusions & Perspectives
Joint PHY-MAC analysis Scientific approach ◮ Physical layer effect on M-CSMA/CA - RTS/CTS PhD Defense Baher Mawlawi November 26, 2015 31 - 63
Joint PHY-MAC analysis Scientific approach ◮ Physical layer effect on M-CSMA/CA - RTS/CTS ◮ Physical layer based on 802.11n standard PhD Defense Baher Mawlawi November 26, 2015 31 - 63
Joint PHY-MAC analysis Scientific approach ◮ Physical layer effect on M-CSMA/CA - RTS/CTS ◮ Physical layer based on 802.11n standard ◮ Performance study according to several MCS values PhD Defense Baher Mawlawi November 26, 2015 31 - 63
Joint PHY-MAC analysis Scientific approach ◮ Physical layer effect on M-CSMA/CA - RTS/CTS ◮ Physical layer based on 802.11n standard ◮ Performance study according to several MCS values ◮ Capture effect PhD Defense Baher Mawlawi November 26, 2015 31 - 63
Joint PHY-MAC analysis Scientific approach ◮ Physical layer effect on M-CSMA/CA - RTS/CTS ◮ Physical layer based on 802.11n standard ◮ Performance study according to several MCS values ◮ Capture effect ◮ Path loss effect considering equal power transmission PhD Defense Baher Mawlawi November 26, 2015 31 - 63
Joint PHY-MAC analysis Scientific approach ◮ Physical layer effect on M-CSMA/CA - RTS/CTS ◮ Physical layer based on 802.11n standard ◮ Performance study according to several MCS values ◮ Capture effect ◮ Path loss effect considering equal power transmission ◮ Performance evaluation according to AWGN and D fading channels PhD Defense Baher Mawlawi November 26, 2015 31 - 63
Joint PHY-MAC analysis Scientific approach ◮ Physical layer effect on M-CSMA/CA - RTS/CTS ◮ Physical layer based on 802.11n standard ◮ Performance study according to several MCS values ◮ Capture effect ◮ Path loss effect considering equal power transmission ◮ Performance evaluation according to AWGN and D fading channels ◮ Successful transmission ratio (STR) for AWGN and D fading channels PhD Defense Baher Mawlawi November 26, 2015 31 - 63
Joint PHY-MAC analysis Scientific approach ◮ Physical layer effect on M-CSMA/CA - RTS/CTS ◮ Physical layer based on 802.11n standard ◮ Performance study according to several MCS values ◮ Capture effect ◮ Path loss effect considering equal power transmission ◮ Performance evaluation according to AWGN and D fading channels ◮ Successful transmission ratio (STR) for AWGN and D fading channels ◮ Interband interference due to asynchronous transmissions PhD Defense Baher Mawlawi November 26, 2015 31 - 63
Joint PHY-MAC analysis Scientific approach ◮ Physical layer effect on M-CSMA/CA - RTS/CTS ◮ Physical layer based on 802.11n standard ◮ Performance study according to several MCS values ◮ Capture effect ◮ Path loss effect considering equal power transmission ◮ Performance evaluation according to AWGN and D fading channels ◮ Successful transmission ratio (STR) for AWGN and D fading channels ◮ Interband interference due to asynchronous transmissions ◮ Analytical model based on OFDM PhD Defense Baher Mawlawi November 26, 2015 31 - 63
Joint PHY-MAC analysis Scientific approach ◮ Physical layer effect on M-CSMA/CA - RTS/CTS ◮ Physical layer based on 802.11n standard ◮ Performance study according to several MCS values ◮ Capture effect ◮ Path loss effect considering equal power transmission ◮ Performance evaluation according to AWGN and D fading channels ◮ Successful transmission ratio (STR) for AWGN and D fading channels ◮ Interband interference due to asynchronous transmissions ◮ Analytical model based on OFDM ◮ SIR analytical expressions considering capture effect PhD Defense Baher Mawlawi November 26, 2015 31 - 63
Joint PHY-MAC analysis Capture effect Scenario PhD Defense Baher Mawlawi November 26, 2015 32 - 63
Joint PHY-MAC analysis Capture effect Scenario 1 R α SIR ( k, i ) = i g k 1 � R α j j =1 j � = i PhD Defense Baher Mawlawi November 26, 2015 32 - 63
Joint PHY-MAC analysis Assumptions ◮ Noise is dominated by interference. PhD Defense Baher Mawlawi November 26, 2015 33 - 63
Joint PHY-MAC analysis Assumptions ◮ Noise is dominated by interference. ◮ The interference is seen like AWGN. PhD Defense Baher Mawlawi November 26, 2015 33 - 63
Recommend
More recommend
