Bhuvan’s status report (Apr 20, 2019)
This week, I worked on creating an application level protocol to transfer packets of data across our network. The protocol works with the receiver first sending a READY packet to the transmitter. The transmitter then sends the data packet to the receiver. The receiver must then send an ACK packet to the transmitter to confirm receipt. If the data is corrupted or no packet is sent after the receiver sends a READY, a NAK is sent to the transmitter indicating that the packet must be resent. Currently I am working on fixing minor issues with the packet formatting as we cannot be sure that all bits in the READY, ACK and NAK packets are sent correctly. I am also working on adding a CRC check to the data packets. I will be working with my teammates to iron out bugs over the weekend.
Dhruva’s Status Report (Apr 13, 2019)
This week, I made significant progress on our capstone. The high-speed op amp came in and I integrated that with the circuit. This required some gain changes as well as incrementally modifying resistor values and supply because the new op amp had different characteristics. These changes along with some changes in the verilog modules allowed us to achieve a speed of 600Kbps which is much much larger than what we had originally intended. I also worked on increasing the distance between the transmitter and the receiver. This required the addition of a switching transistor circuit which I designed and integrated. Adding the transistor increased the distance but only slightly. I will continue to investigate the reasons for such a small increase as well as any other avenues of increasing the seperation between the receiver and transmitted diodes. The in-class demo is on Wednesday next week, and I anticipate that the integration with the application layer will be complete by then.
Raziq’s Status Report (Apr 20, 2019)
This week, I worked with Dhruva on optimizing the modulation/demodulation protocol for speed and distance. This was an iterative process involving testing different constant values. The data size (bits per symbol) has been fixed at 2 since for now we do not need to consider multiple concurrent transmissions to the same receiver. Thus, the pulse width and slot size are the primary factors affecting speed and reliability. We have been able to achieve throughput in around 600 Kbps, with slot size on the order of 800 ns. As far as distance, we are working on increasing the distance beyond 3 cm. This involves both the physical and datalink layer. On my side, the primary way is by adjusting pulse width. Currently, the pulse width cannot be larger than the slot size, but this may be modified as needed.
Upcoming work
Integration and bug fixing for demo
Increase distance between LED and photodiode
Improve speed of data transmission
Team Status Report
We received the high speed op-amps and 9V batteries. We are still waiting on the high speed LEDs to be delivered.