Thursday, November 24, 2005

Cooking For Engineers

Although I prefer eating to cooking, I have dabbled. I came accross a cooking site that makes the though a lot more comfortable: Cooking For Engineers. Theres also an interesting article on Saturated Fats, Cholesterol, and Heart Disease. My conclusions, stop the Potato Chips (Hydrogenated Oils), replace marg with good ol' butter and everything in moderation (I love relativity).

Cooking For Engineers

Although I prefer eating to cooking, I have dabbled. I came accross a cooking site that makes the though a lot more comfortable: Cooking For Engineers. Theres also an interesting article on Saturated Fats, Cholesterol, and Heart Disease. My conclusions, stop the Potato Chips (Hydrogenated Oils), replace marg with good ol' butter and everything in moderation (I love relativity).

Tuesday, October 11, 2005

Engineering News Article

Its amazing how a simple idea like the cantenna has grabbed the attention of the media. So far I've been interviewed on SABC Africa and by Engineering News (Tin-can bridge to digital society). One would have thought that the technical, innovative stuff we're doing with ad-hoc networking in the larger project would have more interest. I suppose that its easier to grasp the imagination with simplicity.

Engineering News Article

Its amazing how a simple idea like the cantenna has grabbed the attention of the media. So far I've been interviewed on SABC Africa and by Engineering News (Tin-can bridge to digital society). One would have thought that the technical, innovative stuff we're doing with ad-hoc networking in the larger project would have more interest. I suppose that its easier to grasp the imagination with simplicity.

Thursday, September 22, 2005

IWS2005 Thursday


IWS2005 Thursday
Keynote speech
WLAN, Fixed Wireless and Cellular all have adopted space-time transmit diversity
Only WLAN’s really use special MUX
Celullar may never see special MUX

Cognitive radio may present a challenge to spatial MUX in WiMAX
MIMO in networks is different from MIMO in isolation

Networks not prepared to cut down on power(loose channels) to include spatial diversity

UWB panel discussion
CSEM – built small, low complexity UWB solution
Low data rate
Tx 3.5 mW
Rx 7.5 mW
Radiated power 100uW
Uses analogue spread spectrum and multi-path diversity
Use of other MAC layer may reduce power further, WiseMAC, S-MAC

CAPANINA Project
Delivering Broadband from a high altitude, quasi stationary plane/airship
120Mb/s symmetric links
Delivery to rural areas, blocks of flats and moving train, using steerable antennae
Optical links for backhaul 48/47GHz 31/28GHz
University of York UK, NICT Japan
60Km diameter user area coverage, single or multiple cell
Possible backhaul via satellite
802.16, 802.20, 802.11 , ETSI Hiperaccess, hiperMAN, DVB
802.16sc is best overall choice
WiMAX, possible depending on regulators
17 – 22 Km above the ground
Dish sizes less than 30 cm
Issues: HAP station keeping and moving train
Antennae: mechatronic and smart
Use 8X8antenna array to block out interfering signals

Test system in Sweden delivered 4Mb/s throughput 802.11b and 1.25Gb/s freespace optic link
Dr. David Grace dg@ohm.york.ac.uk
www.capanina.org
200-300 000 euros for test tethered platform

This has given me an idea!!! Why not put an AP on an advertisement gas balloon as a hop between Jhb and PTA?

Improved Triangulation Method using a Movement model
Localisation done in Bluetooth
Inquiry procedure provides RSSI
Movement model variables: Speed, direction based on previous measurements
Without prior knowledge: Speed truncated normal distribution with a mean of 1m/s and std dev of 0.5m/s
Direction uniform between -180 and 180
With prior knowledge, previous measurements used to determine speed

Planning of Large-scale WLAN infrastructures, Methodius University, Macedonia
Planning either done using educated guesses or with auto planning
Dominant path approach used for propagation modelling
Propogation Modelling, Coverage optimisation and channel alloc optimisation
Wireless Indoor Design Tool, version 1 (WIDT v1) to be released soon
Possible use in indoor mesh?

IWS2005 Thursday


IWS2005 Thursday
Keynote speech
WLAN, Fixed Wireless and Cellular all have adopted space-time transmit diversity
Only WLAN’s really use special MUX
Celullar may never see special MUX

Cognitive radio may present a challenge to spatial MUX in WiMAX
MIMO in networks is different from MIMO in isolation

Networks not prepared to cut down on power(loose channels) to include spatial diversity

UWB panel discussion
CSEM – built small, low complexity UWB solution
Low data rate
Tx 3.5 mW
Rx 7.5 mW
Radiated power 100uW
Uses analogue spread spectrum and multi-path diversity
Use of other MAC layer may reduce power further, WiseMAC, S-MAC

CAPANINA Project
Delivering Broadband from a high altitude, quasi stationary plane/airship
120Mb/s symmetric links
Delivery to rural areas, blocks of flats and moving train, using steerable antennae
Optical links for backhaul 48/47GHz 31/28GHz
University of York UK, NICT Japan
60Km diameter user area coverage, single or multiple cell
Possible backhaul via satellite
802.16, 802.20, 802.11 , ETSI Hiperaccess, hiperMAN, DVB
802.16sc is best overall choice
WiMAX, possible depending on regulators
17 – 22 Km above the ground
Dish sizes less than 30 cm
Issues: HAP station keeping and moving train
Antennae: mechatronic and smart
Use 8X8antenna array to block out interfering signals

Test system in Sweden delivered 4Mb/s throughput 802.11b and 1.25Gb/s freespace optic link
Dr. David Grace dg@ohm.york.ac.uk
www.capanina.org
200-300 000 euros for test tethered platform

This has given me an idea!!! Why not put an AP on an advertisement gas balloon as a hop between Jhb and PTA?

Improved Triangulation Method using a Movement model
Localisation done in Bluetooth
Inquiry procedure provides RSSI
Movement model variables: Speed, direction based on previous measurements
Without prior knowledge: Speed truncated normal distribution with a mean of 1m/s and std dev of 0.5m/s
Direction uniform between -180 and 180
With prior knowledge, previous measurements used to determine speed

Planning of Large-scale WLAN infrastructures, Methodius University, Macedonia
Planning either done using educated guesses or with auto planning
Dominant path approach used for propagation modelling
Propogation Modelling, Coverage optimisation and channel alloc optimisation
Wireless Indoor Design Tool, version 1 (WIDT v1) to be released soon
Possible use in indoor mesh?

Wednesday, September 21, 2005

IWS2005 Wednesday

IWS2005 Thursday
Broadband Rollout and the role of wireless, Steffen Ring, Director Motorola
IEEE802 has 3 categories, Wire line 802.1, 802.3, 802.21 , Wireless 802.11.802.15, 802.16, 802.20 and tech advisory 802.18, 802.19

Enhanced Data rate, 802.11n WLan approx 200+ Mb/s in testing may be able to handle 1000’s of calls and possibly uncompressed video

New mobile extensions to WiMax, 802.16e
Mesh networks 802.11s

802.11a 5GHz 54Mb/s PHY based on OFDM
802.11v 2.4GHz 11Mb/s PHY based on DSSS
802.11d
802.11f Support for roaming
802.11h European 802.11a
802.11i Security
802.11n High Throughput

802.16 Single Carrier MAC/PHY layer 11-66GHz
802.16e Mobile Broadband Wireless Access
WiMax based on 802.16 and 802.16e

http://www.wirelessman.org/tge/index.html

DVB-H as Broadband Access Technology, Reza Tadayoni, CICT, Tech Univ Denmark,
Mobile and Broadcast Convergence
Inband – UMTS and GPRS e.g mibiTV
Hybrid – DVB-H, DMB (Korea), ISDB-T (Japan), MediaFLO (Qualcomm, US)
DVB-H up to 11Mb/s

BREAD project, Broadband access for all in Europe
Mention made of African Cantenna project

Rate Switching in 802.11e, dude from Toshiba
Currently, it is impossible to change the transmission rate during a burst because of Block ACK
4 algorithms proposed to solve the problem
1) link adaptation algorithm for Block ACK
Change and compare ratio of change of rate within a burst
2) Reduce the burst size
3)
4) A tail data packet with a higher rate is attached in every burst

Problem: 802.11 suffers from downlink throughput degradation when uplink traffic increases, cause: AP must transmit all downlink traffic and acquires the channel with the same probability as a single station, therefore to solve the prob, a new scheme is needed.
Solution: In the proposed scheme the AP has a single CSMA/CA function, this lowers prob of collision.
Proposed Scheme compared to DCF and Multiple CSMA/CA
After 26 station Multiple CSMA throughput drops
DCF and proposed scheme seem better, DCF drops off at 50

Current Frame burst handling of frame errors not efficient
Dymanic Queue prioritizing proposed.
Results show that after 30 stations, throughput decreases more in normal round robin than the proposed queue prioritization

IWS2005 Wednesday

IWS2005 Thursday
Broadband Rollout and the role of wireless, Steffen Ring, Director Motorola
IEEE802 has 3 categories, Wire line 802.1, 802.3, 802.21 , Wireless 802.11.802.15, 802.16, 802.20 and tech advisory 802.18, 802.19

Enhanced Data rate, 802.11n WLan approx 200+ Mb/s in testing may be able to handle 1000’s of calls and possibly uncompressed video

New mobile extensions to WiMax, 802.16e
Mesh networks 802.11s

802.11a 5GHz 54Mb/s PHY based on OFDM
802.11v 2.4GHz 11Mb/s PHY based on DSSS
802.11d
802.11f Support for roaming
802.11h European 802.11a
802.11i Security
802.11n High Throughput

802.16 Single Carrier MAC/PHY layer 11-66GHz
802.16e Mobile Broadband Wireless Access
WiMax based on 802.16 and 802.16e

http://www.wirelessman.org/tge/index.html

DVB-H as Broadband Access Technology, Reza Tadayoni, CICT, Tech Univ Denmark,
Mobile and Broadcast Convergence
Inband – UMTS and GPRS e.g mibiTV
Hybrid – DVB-H, DMB (Korea), ISDB-T (Japan), MediaFLO (Qualcomm, US)
DVB-H up to 11Mb/s

BREAD project, Broadband access for all in Europe
Mention made of African Cantenna project

Rate Switching in 802.11e, dude from Toshiba
Currently, it is impossible to change the transmission rate during a burst because of Block ACK
4 algorithms proposed to solve the problem
1) link adaptation algorithm for Block ACK
Change and compare ratio of change of rate within a burst
2) Reduce the burst size
3)
4) A tail data packet with a higher rate is attached in every burst

Problem: 802.11 suffers from downlink throughput degradation when uplink traffic increases, cause: AP must transmit all downlink traffic and acquires the channel with the same probability as a single station, therefore to solve the prob, a new scheme is needed.
Solution: In the proposed scheme the AP has a single CSMA/CA function, this lowers prob of collision.
Proposed Scheme compared to DCF and Multiple CSMA/CA
After 26 station Multiple CSMA throughput drops
DCF and proposed scheme seem better, DCF drops off at 50

Current Frame burst handling of frame errors not efficient
Dymanic Queue prioritizing proposed.
Results show that after 30 stations, throughput decreases more in normal round robin than the proposed queue prioritization

IWS2005 Wednesday Afternoon

Power Efficient
Wireless is unrelieable, high BER
Leads to high power consumption
Use Co-operative Diversity (spatial diversity)
Does this lower transmission power of individual nodes?
REalisation:
hardware - Mulitple antenna on every node, minimum distance between antennas should be l/2 = 6.12 cm
software - Space -time codes, two dimensional array
Nodes with single antennas cop-operate to create a virual MIMO, which gives the same effect [Laneman et al]

IWS2005 Wednesday Afternoon

Power Efficient
Wireless is unrelieable, high BER
Leads to high power consumption
Use Co-operative Diversity (spatial diversity)
Does this lower transmission power of individual nodes?
REalisation:
hardware - Mulitple antenna on every node, minimum distance between antennas should be l/2 = 6.12 cm
software - Space -time codes, two dimensional array
Nodes with single antennas cop-operate to create a virual MIMO, which gives the same effect [Laneman et al]

Tuesday, September 20, 2005

IWS 2005 Tuesday

Home Networking: Sudhir Dixit, Research Fellow, Nokia Research Center
Home Networking; the next frontier (not the final frontier, coz thats been reserved for space)

DLNA consortium
DLNA Vision, promote interoperaability between home networked devices.

Media Exchange protocols for home networks, Edwin Heredia, Microsoft

Location based services, Paul Wisner, Nokia
Based on Nokia 6630, uses Discovery Protocols: Bluetooth SDP and UPnP
Others: SLP and JINI
Nokia Approach: Everything is aware of its own location, proximity is determined by comparisn
Map location to position, position determined from GPS, WLAN triangulation etc.
Desktop on mobile device shows reachable services

Usable Security in Smart Homes, Saad Shakshir, Nokia Research


Hot Topics: MIMO
What is MIMO?
Definition: MIMO

IWS 2005 Tuesday

Home Networking: Sudhir Dixit, Research Fellow, Nokia Research Center
Home Networking; the next frontier (not the final frontier, coz thats been reserved for space)

DLNA consortium
DLNA Vision, promote interoperaability between home networked devices.

Media Exchange protocols for home networks, Edwin Heredia, Microsoft

Location based services, Paul Wisner, Nokia
Based on Nokia 6630, uses Discovery Protocols: Bluetooth SDP and UPnP
Others: SLP and JINI
Nokia Approach: Everything is aware of its own location, proximity is determined by comparisn
Map location to position, position determined from GPS, WLAN triangulation etc.
Desktop on mobile device shows reachable services

Usable Security in Smart Homes, Saad Shakshir, Nokia Research


Hot Topics: MIMO
What is MIMO?
Definition: MIMO

IWS2005 Tues Afternoon

Duplicate Address Detection in OLSR networks, Thomas Clausen, Inria

MANETS: Mobile Adhoc Networks, do not comply with traditional internet hypothesis
Networks split and re-combine and may cause address conflicts.
DHCP is statefull , server based and sensitive to partitioning
IPV6 also has similar problems
Pro-active protocol
Inspired by NOA OLSR and PACMAN
The proposed solution is to use link state messages to determine the occurance of duplicate addresses and initiate self address change in case of detection

AD-hoc network applcations and technical issues, Dude from Hitachi
3 uses for ad-hoc:
specialized, e.g miltary, emergency, disaster
enterprise e.g. office, home
consumer e.g. community
Town security System with wireless cameras, Tracking, Cellular phones using ad-hoc networking to bypass network
ITS service systems
3 test systems
Security system with 8 olsr nodes, cameras and pda
Car mobile network with auto handover between wilan and cellular, using olsr and mobile IPv6

Future of Wireless

Future of Wireless
MIMO
Cognitive Radio, implemented using software radio, to improve freq utilization
Ubiquitous networking, networking everything (man and machine)
RFID, sensor networks

IWS2005 Tues Afternoon

Duplicate Address Detection in OLSR networks, Thomas Clausen, Inria

MANETS: Mobile Adhoc Networks, do not comply with traditional internet hypothesis
Networks split and re-combine and may cause address conflicts.
DHCP is statefull , server based and sensitive to partitioning
IPV6 also has similar problems
Pro-active protocol
Inspired by NOA OLSR and PACMAN
The proposed solution is to use link state messages to determine the occurance of duplicate addresses and initiate self address change in case of detection

AD-hoc network applcations and technical issues, Dude from Hitachi
3 uses for ad-hoc:
specialized, e.g miltary, emergency, disaster
enterprise e.g. office, home
consumer e.g. community
Town security System with wireless cameras, Tracking, Cellular phones using ad-hoc networking to bypass network
ITS service systems
3 test systems
Security system with 8 olsr nodes, cameras and pda
Car mobile network with auto handover between wilan and cellular, using olsr and mobile IPv6

Future of Wireless

Future of Wireless
MIMO
Cognitive Radio, implemented using software radio, to improve freq utilization
Ubiquitous networking, networking everything (man and machine)
RFID, sensor networks

Monday, September 19, 2005

IWS 2005 Monday Afternoon notes

MOCCA: Summary of think-tank
Focus is on Asian developing market, particularly India and China.
Debate between whether investment into existing technology in rural sector of mobile market will have better returns than further urban investment into new technology.
Debate on whether voice or data services are more important in rural areas.
Debate on what a rural area is? Low density population?

Mobile Handover Strategies, A. Slingerland
Mobility management scheme supports multiple strategies and smartly selects best strategy which will improve QoS
Application requirements; latency, loss, re-ordering
Context issues; link layer hints, network support nodes, load
Hierarchical Mobile IP
Static handover is sub-optimal

Distributed QoS control Arch for latency sensitive applications, Mahbubul Alam
QoS, multi-layer challenge
Basic Challengers: BW, Delay and Jitter
Arch challenges: Scalabity, Resource management, Robustness against traffic fluctuations, Billing, Security
Seamless security needs to be included in design along with QoS.
Security can provide QoS, e.g. ipsec, however not scalable
QoS routing policy

M. Canales (mcanales@unizar.es)
Shortest path not necessarily QoS, resource utilisation at MAC layer more important
New QoS metric based on Mac resources
Distruitbuted
Basis: AOMDV
ADHOC MAC protocol
Destination selects best path after receiving multipath info / metric
Simulation: Event driven simulator in c++ ( Adhoc MAC + modified aomdv+path bandwidth calc algorithm)
Best effort strategy: AODV
QoS metric based strategy QoS Routing, AOMDV

Distributed Admission Control Algorithm
IEEE 802.11e EDCA QoS Concept

QoS relate to RT and non-Rt services, which may have contrasting delays and reliability requirements, flexible QoS management required.
Proposed system based on ITU-R
Vertical and horizontal handover

IWS 2005 Monday Afternoon notes

MOCCA: Summary of think-tank
Focus is on Asian developing market, particularly India and China.
Debate between whether investment into existing technology in rural sector of mobile market will have better returns than further urban investment into new technology.
Debate on whether voice or data services are more important in rural areas.
Debate on what a rural area is? Low density population?

Mobile Handover Strategies, A. Slingerland
Mobility management scheme supports multiple strategies and smartly selects best strategy which will improve QoS
Application requirements; latency, loss, re-ordering
Context issues; link layer hints, network support nodes, load
Hierarchical Mobile IP
Static handover is sub-optimal

Distributed QoS control Arch for latency sensitive applications, Mahbubul Alam
QoS, multi-layer challenge
Basic Challengers: BW, Delay and Jitter
Arch challenges: Scalabity, Resource management, Robustness against traffic fluctuations, Billing, Security
Seamless security needs to be included in design along with QoS.
Security can provide QoS, e.g. ipsec, however not scalable
QoS routing policy

M. Canales (mcanales@unizar.es)
Shortest path not necessarily QoS, resource utilisation at MAC layer more important
New QoS metric based on Mac resources
Distruitbuted
Basis: AOMDV
ADHOC MAC protocol
Destination selects best path after receiving multipath info / metric
Simulation: Event driven simulator in c++ ( Adhoc MAC + modified aomdv+path bandwidth calc algorithm)
Best effort strategy: AODV
QoS metric based strategy QoS Routing, AOMDV

Distributed Admission Control Algorithm
IEEE 802.11e EDCA QoS Concept

QoS relate to RT and non-Rt services, which may have contrasting delays and reliability requirements, flexible QoS management required.
Proposed system based on ITU-R
Vertical and horizontal handover

Notes taken: Exec Round Table Discussion

Dude from Siemens:
Performance increase by 100X every 5 yrs
Experimentation currently at 1Gbs

Higher frequencies aimed at for 4g have less coverage.

Dude from board of every 4 letter acronym known to man:
(KDDI R&D) Dr. Yasuo Hirate

Japan uses W-CDMA (90 mil users)

Aimed at 100Mbs by 2010

Fixes mobile convergence
Seamless network of mobile and wireless lan
Integration, Seamless, Barrier Free
Convergence of Fixed, Wireless, Cellular and Digital Broadcast

Dude from Qualcomm, formerly AT&T Bell Labs:
Anil Kripalani
Ubiquitous Search and Discovery

Coming soon: Phones with hard-disks and 7.1 Mega pixel camera

Dude from Nokia:
Dr. Jan Bosch
“Lifepod” – Record everything

Dude from Saumsung:
Dr. Young Kyun Kim
Telecoms and Broadcasting convergence
4g : IP Ubiquitous Network
Mobile terminal as gateway to WiLan, BT, RFID, IrDA and IP Network
IPv6 based network
2 paths to 4G, 1 is the cellular path 1G, 2G, 3G ( , 2nd path is the “nomadic” path, wlan, BT, Zigbee, RFID ( 4G
WiBro (Mobile WiMax), ubiquitous networks in Korea
Apec summit in Nusan Korea will demo large volume VoD, MoD

Dude from Intel:
Dan Dahle
3g for WAN
WiMax for MAN
WiFi for LAN
UWB and BT fot PAN
NFC, RFID
Wireless Requirement
Throughput, Latency, tolerance for loss, mobility, interference, capacity, bandwidth, throughput, power consumption, complexity

Dudette from Alcatel:
Simplicity, continuum of service
Heterogeneous Access Management required for multiple radio systems.
Multi-radio terminal, transparent to user and power-wise
IMS architecture enabler of new applications


Summary
4G is a network of networks, it is a solution looking for a problem.
Terminal has to accommodate converged radio networks.
“The user wants to google cnn anytime, anywhere at the right cost having all security considered…”

Notes taken: Exec Round Table Discussion

Dude from Siemens:
Performance increase by 100X every 5 yrs
Experimentation currently at 1Gbs

Higher frequencies aimed at for 4g have less coverage.

Dude from board of every 4 letter acronym known to man:
(KDDI R&D) Dr. Yasuo Hirate

Japan uses W-CDMA (90 mil users)

Aimed at 100Mbs by 2010

Fixes mobile convergence
Seamless network of mobile and wireless lan
Integration, Seamless, Barrier Free
Convergence of Fixed, Wireless, Cellular and Digital Broadcast

Dude from Qualcomm, formerly AT&T Bell Labs:
Anil Kripalani
Ubiquitous Search and Discovery

Coming soon: Phones with hard-disks and 7.1 Mega pixel camera

Dude from Nokia:
Dr. Jan Bosch
“Lifepod” – Record everything

Dude from Saumsung:
Dr. Young Kyun Kim
Telecoms and Broadcasting convergence
4g : IP Ubiquitous Network
Mobile terminal as gateway to WiLan, BT, RFID, IrDA and IP Network
IPv6 based network
2 paths to 4G, 1 is the cellular path 1G, 2G, 3G ( , 2nd path is the “nomadic” path, wlan, BT, Zigbee, RFID ( 4G
WiBro (Mobile WiMax), ubiquitous networks in Korea
Apec summit in Nusan Korea will demo large volume VoD, MoD

Dude from Intel:
Dan Dahle
3g for WAN
WiMax for MAN
WiFi for LAN
UWB and BT fot PAN
NFC, RFID
Wireless Requirement
Throughput, Latency, tolerance for loss, mobility, interference, capacity, bandwidth, throughput, power consumption, complexity

Dudette from Alcatel:
Simplicity, continuum of service
Heterogeneous Access Management required for multiple radio systems.
Multi-radio terminal, transparent to user and power-wise
IMS architecture enabler of new applications


Summary
4G is a network of networks, it is a solution looking for a problem.
Terminal has to accommodate converged radio networks.
“The user wants to google cnn anytime, anywhere at the right cost having all security considered…”

International Wireless Summit 2005

International Wireless Summit 2005, Aalborg, Denmark

The international wireless summit has just kicked off and I have the privilege of being amongst some of the greatest think-tanks in the wireless arena.

The aim of the IWS is to offer a platform for establishing exchanges of information between universities, industry and science parks. The next summit will be in 2008 in Helsinki, Finland.

International Wireless Symposium aims to exchange research information. 450 papers to be presented on “hot topics”.



Prof. Ramjee Prasad giving introductory speech.

Prof. Ramjee Prasad predicts that the future of wireless comms will be based on single layer technology, not quite sure what is meant by this though.

“Personal Networks are going to be of major importance in future business…”

International Wireless Summit 2005

International Wireless Summit 2005, Aalborg, Denmark

The international wireless summit has just kicked off and I have the privilege of being amongst some of the greatest think-tanks in the wireless arena.

The aim of the IWS is to offer a platform for establishing exchanges of information between universities, industry and science parks. The next summit will be in 2008 in Helsinki, Finland.

International Wireless Symposium aims to exchange research information. 450 papers to be presented on “hot topics”.



Prof. Ramjee Prasad giving introductory speech.

Prof. Ramjee Prasad predicts that the future of wireless comms will be based on single layer technology, not quite sure what is meant by this though.

“Personal Networks are going to be of major importance in future business…”

Friday, September 09, 2005

The next Big thing....

Mobile companies have spotted it a while ago and its already filtered through to service providers in the form of "look-for-me" type services. But one company has just launched location based services and perhaps the Internet to a whole new level; Hello Google Earth. Its been a while since anything in the pc/cyber world has slapped a wow at me and it was worth the wait! The uses are already incredible and the possibilities are amazing. Imagine browsing the net by finding pages based on where they're physically kept, like spinning the globe to Google headquartes to launch a google search?

Have to interrupt this transmission, but will resume later...

The next Big thing....

Mobile companies have spotted it a while ago and its already filtered through to service providers in the form of "look-for-me" type services. But one company has just launched location based services and perhaps the Internet to a whole new level; Hello Google Earth. Its been a while since anything in the pc/cyber world has slapped a wow at me and it was worth the wait! The uses are already incredible and the possibilities are amazing. Imagine browsing the net by finding pages based on where they're physically kept, like spinning the globe to Google headquartes to launch a google search?

Have to interrupt this transmission, but will resume later...

Network stats on Freifunk

See: http://wiki.freifunk-leipzig.public-ip.org/index.php/LinksysNetzwerkStatisik for info on how to install web interface packages for network stats.
p.s. hows your german?

Network stats on Freifunk

See: http://wiki.freifunk-leipzig.public-ip.org/index.php/LinksysNetzwerkStatisik for info on how to install web interface packages for network stats.
p.s. hows your german?

Tuesday, August 30, 2005

Its ALIVE!!!!

In just a few weeks the mesh has grown considerably and will hopefully continue to do so. Viva le revolution!!!

Its ALIVE!!!!

In just a few weeks the mesh has grown considerably and will hopefully continue to do so. Viva le revolution!!!

Monday, August 01, 2005

wild_west_mesh

One of the projects that are being undertaken at work that has caught my personal interest is the "Community Owned Information Networks" project(See Project Website and COIN Blog). The aim of the project is to research the use of wireless communication technology, usually used in typical 1st world environments such as offices or neighbourhood networks, in a 3rd world, African context. Networks have been set up in rural villages, linking clinics to hospitals and encouraging rural communities to connect themselves to the networks, by demonstrating the use of alternative, cheaper equipment such as cantenna's. As part of the research an outdoor mesh network was setup in Pretoria and findings form this network are continuously passed back to the rural implementations. Furthermore, a 50 node, indoor network is being setup to further encourage research into mesh networking, including protocol level research.

Although I'm not currently active in the project, I have been keeping up with developments and while scanning the activity around Mesh networking in South Africa, came accross the Johannesburg Areas Wireless users Group, JAWUG.I thus saw an opportunity to experiment with the technology and join the revolution by setting up my own node as well as an opportunity to contribute findings between the CSIR and JAWUG networks. Its a real pity that there isnt any communication/collaboration between the two groups, because of somewhat irresponsible/ childish? ramblings of a core JAWUG member, all of whose critisms are not totally unfounded, but are mostly uninformed. The fact is that the Meraka insitute is only a few months old, taking its name from the CSIR's Open Source Center, which was merged into the new institute. Meraka's interests incorporate ICT's in general, of which connectivity is one theme in pool of many including Human Language Technology, Advanced Computing and Accessibilty. The research into mesh networking is one sub-project within the Wireless Africa project, which itself falls into the connectivty theme.

Afer communicating with nearby interested parties, which I found through the JAWUG and NodeDB websites , buying and setting up equipment, I finally found myself connected...

The following is a network diagram, generated by a routing protocol (OLSR) used on my mesh node.

wild_west_mesh

One of the projects that are being undertaken at work that has caught my personal interest is the "Community Owned Information Networks" project(See Project Website and COIN Blog). The aim of the project is to research the use of wireless communication technology, usually used in typical 1st world environments such as offices or neighbourhood networks, in a 3rd world, African context. Networks have been set up in rural villages, linking clinics to hospitals and encouraging rural communities to connect themselves to the networks, by demonstrating the use of alternative, cheaper equipment such as cantenna's. As part of the research an outdoor mesh network was setup in Pretoria and findings form this network are continuously passed back to the rural implementations. Furthermore, a 50 node, indoor network is being setup to further encourage research into mesh networking, including protocol level research.

Although I'm not currently active in the project, I have been keeping up with developments and while scanning the activity around Mesh networking in South Africa, came accross the Johannesburg Areas Wireless users Group, JAWUG.I thus saw an opportunity to experiment with the technology and join the revolution by setting up my own node as well as an opportunity to contribute findings between the CSIR and JAWUG networks. Its a real pity that there isnt any communication/collaboration between the two groups, because of somewhat irresponsible/ childish? ramblings of a core JAWUG member, all of whose critisms are not totally unfounded, but are mostly uninformed. The fact is that the Meraka insitute is only a few months old, taking its name from the CSIR's Open Source Center, which was merged into the new institute. Meraka's interests incorporate ICT's in general, of which connectivity is one theme in pool of many including Human Language Technology, Advanced Computing and Accessibilty. The research into mesh networking is one sub-project within the Wireless Africa project, which itself falls into the connectivty theme.

Afer communicating with nearby interested parties, which I found through the JAWUG and NodeDB websites , buying and setting up equipment, I finally found myself connected...

The following is a network diagram, generated by a routing protocol (OLSR) used on my mesh node.

Thursday, July 14, 2005

Wikipedia

Created an entry for myself on Wikipedia, dont ask me why though...., excellant source of info though. Also added some bits on Rustenburg.

Wikipedia

Created an entry for myself on Wikipedia, dont ask me why though...., excellant source of info though. Also added some bits on Rustenburg.

Friday, April 29, 2005

The new home of my Ubuntu Linux related blogs: yusufk's Journal

I've decided to move my Ubuntu related blogs to Ubuntu Linux Forums - vB Journal: yusufk's Journal, beacuse its closer to where I usually find answers anyway. This blog maybe quieter though, not that anyone gives a nut, coz I'm the only one that reads this, hehe.

How to turn on Num Lock on GNOME startup?

Unofficial Ubuntu 5.04 Starter Guide: "


sudo apt-get install numlockx

sudo cp /etc/X11/gdm/Init/Default /etc/X11/gdm/Init/Default_backup

sudo gedit /etc/X11/gdm/Init/Default


4. Find this line


...

exit 0


5. Add the following lines above it


if [ -x /usr/bin/numlockx ]; then

/usr/bin/numlockx on

fi


6. Save the edited file (sample)

7. Read How to restart GNOME without rebooting computer?"

The new home of my Ubuntu Linux related blogs: yusufk's Journal

I've decided to move my Ubuntu related blogs to Ubuntu Linux Forums - vB Journal: yusufk's Journal, beacuse its closer to where I usually find answers anyway. This blog maybe quieter though, not that anyone gives a nut, coz I'm the only one that reads this, hehe.

How to turn on Num Lock on GNOME startup?

Unofficial Ubuntu 5.04 Starter Guide: "

sudo apt-get install numlockx
sudo cp /etc/X11/gdm/Init/Default /etc/X11/gdm/Init/Default_backup
sudo gedit /etc/X11/gdm/Init/Default

4. Find this line

...
exit 0

5. Add the following lines above it

if [ -x /usr/bin/numlockx ]; then
/usr/bin/numlockx on
fi

6. Save the edited file (sample)
7. Read How to restart GNOME without rebooting computer?"

Thursday, April 21, 2005

How to make Ubuntu/Debian .deb packages from source tar.gz

This is my abbrevitated version of : Hands On


Step 1: Make the source files

Un-tar your source files. This should result in a directory ./the_app

cd into the_app directory: cd the_app

Make the application according to the instructions in the README file, usually just means running the make command in the the_app directory: ./make


Step 2: Prepare the "control" file

Then create a directory DEBIAN: mkdir DEBIAN

cd into DEBIAN: cd DEBIAN

Create a file called "control" based on the following example:


Package: your_package_name_without_ext

Version: 1.0

Section: base

Priority: optional

Architecture: all

Depends: prog1 (>= 2.05a-11), prog2 (>= 2.0-12), prog3, prog4 (>= 1:2.0.7-8), prog5 (>= 3.02-8), prog6 (>= 2.4.2-3), prog7 (>= 5.0-5)

Maintainer: Yusuf

Description: A little sumthing sumthing on your program


Step 3: Put all the files in their places

Now you've got to create a directory structure with the files needed as they will be installed on your system.

So if you need the bin files in your /usr/local/bin directory then copy the bin files into the_app/usr/local/bin : cp the_app.sh usr/local/bin


Step 4: Create the Package

Decend into the directory containing the_app and run: dpkg-deb --build the_app

Thats it! You should now have a the_app.deb package


Extract a gzip compressed tar archive in Linux

To extract the archive filename.tar.gz into the current directory:


tar xzf filename.tar.gz


If this fails, the version of tar may not support gzip compression. In this case, you can use the traditional two-stage command:


gzip -dc filename.tar.gz | tar xf -

How to make Ubuntu/Debian .deb packages from source tar.gz

This is my abbrevitated version of : Hands On

Step 1: Make the source files
Un-tar your source files. This should result in a directory ./the_app
cd into the_app directory: cd the_app
Make the application according to the instructions in the README file, usually just means running the make command in the the_app directory: ./make

Step 2: Prepare the "control" file
Then create a directory DEBIAN: mkdir DEBIAN
cd into DEBIAN: cd DEBIAN
Create a file called "control" based on the following example:

Package: your_package_name_without_ext
Version: 1.0
Section: base
Priority: optional
Architecture: all
Depends: prog1 (>= 2.05a-11), prog2 (>= 2.0-12), prog3, prog4 (>= \ 1:2.0.7-8), prog5 (>= 3.02-8), prog6 (>= 2.4.2-3), prog7 (>= 5.0-5)
Maintainer: Yusuf
Description: A little sumthing sumthing on your program

Step 3: Put all the files in their places
Now you've got to create a directory structure with the files needed as they will be installed on your system.
So if you need the bin files in your /usr/local/bin directory then copy the bin files into the_app/usr/local/bin : cp the_app.sh usr/local/bin

Step 4: Create the Package
Decend into the directory containing the_app and run: dpkg-deb --build the_app
Thats it! You should now have a the_app.deb package

Extract a gzip compressed tar archive in Linux

To extract the archive filename.tar.gz into the current directory:

tar xzf filename.tar.gz

If this fails, the version of tar may not support gzip compression. In this case, you can use the traditional two-stage command:

gzip -dc filename.tar.gz | tar xf -

Tuesday, April 19, 2005

APT syncing a home machine

Thanks to Hellkom, its not worth it for me (or any South African for that matter) to get an ADSL line at home. I won't even mention dialup (ok I have, but I regret it). So this creates difficulties. One of them is the fact that it is a pain to keep my home maching up to date with the latest ubuntu packages.

I sought advice and I found it (courtesy Thomas Fogwill):

There are at least 3 other ways to do what you want:
1) Repository on notebook
2) Manually generate fetch lists
3) Use automated tools

(1) Repository on notebook (I do this)
I keep my downloaded packages on my notebook
(in /var/cache/apt/archives/). Things I want to install at home, but not
on my notebook, are downloaded with apt-get -d install ...
I have a www directory linked to /var/cache/apt/archives/,
and generate a Packages.gz file here like this:
apt-ftparchive packages . | gzip -c > Packages.gz

Then, I add this www directory to my wife's sources.list (at home). If
any packages are ever missing from this dir, I rebuild them with
dpkg-repack.

The following is a useful little script to rebuild ALL installed
packages:
dpkg --get-selections | grep [^De]install | cut -f1 | sh -c "while read STRIN; do dpkg-repack $STRIN ;done"

Note: dpkg-repack uses the files on the filesystem to rebuild the
package. Thus, any config/customisation you've done will be included in
the deb.

2) Manually generate fetch lists
This should also work (untested). Apt stores its downloaded package
files here: /var/lib/apt/lists/
You should be able to copy your updated files over from your notebook,
then do an apt-get --print-uris install|upgrade|dist-upgrade on the home
machine. This will print out the uri's of the files that need to be
downloaded. You then take these uri's, download them (at work) with
wget, copy them to the home machine and dpkg -i *.deb to install them.

This will dump the apt-get output into a file that wget can use to
download in batch mode:
apt-get --print-uris -y install ... | grep http | cut -f1 -d' ' | cut -f2 -d' > debs.toget
To batch fetch the debs:
wget -i debs.toget

This should work fairly well, except when packages on the mirror change overnight.

3) Use automated tools
e.g. apt-zip. apt-zip basically automates (2) above.

APT syncing a home machine

Thanks to Hellkom, its not worth it for me (or any South African for that matter) to get an ADSL line at home. I won't even mention dialup (ok I have, but I regret it). So this creates difficulties. One of them is the fact that it is a pain to keep my home maching up to date with the latest ubuntu packages.

I sought advice and I found it (courtesy Thomas Fogwill):

There are at least 3 other ways to do what you want:
1) Repository on notebook
2) Manually generate fetch lists
3) Use automated tools

(1) Repository on notebook (I do this)
I keep my downloaded packages on my notebook
(in /var/cache/apt/archives/). Things I want to install at home, but not
on my notebook, are downloaded with apt-get -d install ...
I have a www directory linked to /var/cache/apt/archives/,
and generate a Packages.gz file here like this:
apt-ftparchive packages . | gzip -c > Packages.gz

Then, I add this www directory to my wife's sources.list (at home). If
any packages are ever missing from this dir, I rebuild them with
dpkg-repack.

The following is a useful little script to rebuild ALL installed
packages:
dpkg --get-selections | grep [^De]install | cut -f1 | sh -c "while read STRIN; do dpkg-repack \$STRIN ;done"

Note: dpkg-repack uses the files on the filesystem to rebuild the
package. Thus, any config/customisation you've done will be included in
the deb.

2) Manually generate fetch lists
This should also work (untested). Apt stores its downloaded package
files here: /var/lib/apt/lists/
You should be able to copy your updated files over from your notebook,
then do an apt-get --print-uris install|upgrade|dist-upgrade on the home
machine. This will print out the uri's of the files that need to be
downloaded. You then take these uri's, download them (at work) with
wget, copy them to the home machine and dpkg -i *.deb to install them.

This will dump the apt-get output into a file that wget can use to
download in batch mode:
apt-get --print-uris -y install ... | grep http | cut -f1 -d' ' | cut -f2 -d\' > debs.toget
To batch fetch the debs:
wget -i debs.toget

This should work fairly well, except when packages on the mirror change overnight.

3) Use automated tools
e.g. apt-zip. apt-zip basically automates (2) above.

Saturday, April 16, 2005

Ubuntu - Linux for Human Beings

My 1st experience with a pc was with a ZX Spectrum way back in 1990'ish. I later progressed to a Commodore 128 and then finally in 1992 I mananged to get my paws on on my dads 286, which he intended to use for bussiness. DOS was a dream, it was complex enough (at the time) to keep me challenged for hours on end, especially when I had recover (on more than one occasion) from the odd virus, DIR II, Michaelangelo, Stoned, the good old days. When I discovered Windows in 1995, it really was not my forte, it hid the complexity of the pc too much and made things too easy. But, I had to stick with the times and the Games, which soon required Windows and Windows soon lost DOS.


To cut a long story short, for now, until I get to update this, I discovered Ubuntu Linux, much better than my previous encounters with Linux (Mandrake, Red-Hat), works like a charm and its FREE! Check out: Ubuntu - Linux for Human Beings

Ubuntu - Linux for Human Beings

My 1st experience with a pc was with a ZX Spectrum way back in 1990'ish. I later progressed to a Commodore 128 and then finally in 1992 I mananged to get my paws on on my dads 286, which he intended to use for bussiness. DOS was a dream, it was complex enough (at the time) to keep me challenged for hours on end, especially when I had recover (on more than one occasion) from the odd virus, DIR II, Michaelangelo, Stoned, the good old days. When I discovered Windows in 1995, it really was not my forte, it hid the complexity of the pc too much and made things too easy. But, I had to stick with the times and the Games, which soon required Windows and Windows soon lost DOS.

To cut a long story short, for now, until I get to update this, I discovered Ubuntu Linux, much better than my previous encounters with Linux (Mandrake, Red-Hat), works like a charm and its FREE! Check out: Ubuntu - Linux for Human Beings

Tuesday, March 15, 2005

George Dubya vs Einstein

RmEinstein was a genius....

"He who joyfully marches to music in rank and file has already earned my contempt. He has been given a large brain by mistake, since for him the spinal cord would fully suffice. This disgrace to civilization should be done away with at once. Heroism at command, senseless brutality, deplorable love-of-country stance, how violently I hate all this, how despiceable an ignoreable war is; I would rather be torn to shreds than be a part of so base an action! It is my conviction that killing under the cloak of war is nothing but an act of murder." -- Albert Einstein

"You cannot simultaneously prevent and prepare for war." --Albert Einstein (1879-1955)

Dubya is not....

"The reason we start a war is to fight a war, win a war, thereby causing no more war!"
--The first Presidential debate

"The war on terror involves Saddam Hussein because of the nature of Saddam Hussein, the history of Saddam Hussein, and his willingness to terrorize himself."-Grand Rapids, Mich., Jan. 29, 2003

George Dubya vs Einstein

RmEinstein was a genius....

"He who joyfully marches to music in rank and file has already earned my contempt. He has been given a large brain by mistake, since for him the spinal cord would fully suffice. This disgrace to civilization should be done away with at once. Heroism at command, senseless brutality, deplorable love-of-country stance, how violently I hate all this, how despiceable an ignoreable war is; I would rather be torn to shreds than be a part of so base an action! It is my conviction that killing under the cloak of war is nothing but an act of murder." -- Albert Einstein

"You cannot simultaneously prevent and prepare for war." --Albert Einstein (1879-1955)

Dubya is not....

"The reason we start a war is to fight a war, win a war, thereby causing no more war!"
--The first Presidential debate

"The war on terror involves Saddam Hussein because of the nature of Saddam Hussein, the history of Saddam Hussein, and his willingness to terrorize himself."-Grand Rapids, Mich., Jan. 29, 2003

Friday, February 25, 2005

Online entry form for the Young Science Writer's Competition

Yep, I considered entering the competition Online entry form for the Young Science Writer's Competition ,couldnt decide on a topic so I though I'd explain EVERYTHING, that is Life the Universe and everything..

Heres my 1st paragraph:

Its the question that we've all asked ourselves at some point, what is Life? But how many have started answering the question by first asking, what is a point? A point is a position in time and an instance in time, two of the three defining axis of our Universe. Energy, the third axis is what everything is made of. You, your peanut butter sandwich and the 3rd left udder of a mountain goat on Table Mountain are energy in its slowest form, matter. On the other hand, Egoli on its way to your TV, the orange warmth of sunrise on its way from the sun and the flux that heats your microwave noodles are energy in its fastest form, electro-magnetic waves.

Online entry form for the Young Science Writer's Competition

Yep, I considered entering the competition Online entry form for the Young Science Writer's Competition ,couldnt decide on a topic so I though I'd explain EVERYTHING, that is Life the Universe and everything..

Heres my 1st paragraph:

Its the question that we've all asked ourselves at some point, what is Life? But how many have started answering the question by first asking, what is a point? A point is a position in time and an instance in time, two of the three defining axis of our Universe. Energy, the third axis is what everything is made of. You, your peanut butter sandwich and the 3rd left udder of a mountain goat on Table Mountain are energy in its slowest form, matter. On the other hand, Egoli on its way to your TV, the orange warmth of sunrise on its way from the sun and the flux that heats your microwave noodles are energy in its fastest form, electro-magnetic waves.