Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

[RPRWG] some questions




Dear Leaders,

    After reading IEEE 802.17 Gandalf Proposal, I feel confused about some part. So I'd like to ask some stupid questions.
    1. In Gandalf page 35, it says that access control can be broken into Global access control and Local access control... I don't understand well about Global access control and I was even confused by the example. In the example refered to Figure27, it says that nodes 1,2,5 will get 1/2 of the bandwidth on a global allocation basis. Why? In my understanding of Global access control, each of which should get 1/3 of the bandwidth, because there are 3 flows in the outer ring. So can you tell me what's wrong with my understanding?
    2. At page 37, it says that Medium proority traffic is asigned a CAR and traffic within the CAR is treated as Hi-pri traffic and traffic above the CAR will be treated as low-pri, refered to eMP. I really don't know the exact meaning about CAR, why we talk about this thing. In RFC2892, there are only two class. Is it not enough?
    
    There are more foolish question about RPR-fa, I'd be deeply appreciative if you will go on to read.
    3. At page 42, it says that each node is assigned a weight which allows the user to allocate more ring bandwidth to certain nodes. What's the function of "Weight"? Wy should we define such a variable?
    4. At page 42, if a node experiences congestion, it will advertise the value of its transmit fiar rate counter(my_rate) to upstream nodes via the opposite ring. The fair rate counter is run through a low pass filter function and divided by the node's weight... I'm really confused by the term low pass filter, I always think of the low pass filter in signal processing. How does this low pass filter work? Would you give an example?
    5. At page 43, I don't understand well about traffic policing function, would you explain it by example? 
    6. At page 44, a node that receives a non-null fairness message will set its allow_rate_congestion and TTL_to_congestion to the rcvd_rate value multiplied by its weight and 255-received TTL value. This allows a node with a weight of N to utilize N times as much bandwidth as a node with a weight of 1.0. I really don't understand the meaning of this part... what's the idea of it?
    7. About pseudo-code, why we define Constants such values
                            Uint16 AGECOEFF=4;
                            Uint16 LP_FWD=64;
                            Uint16 LP_MU=512;
                            Uint16 LP_ALLOW=64;
                            Uint16 LP_ALLOW_COEF=128;
    8. What is MAX_LRATE? Is it equal to link capacity?
    9. What is DECAY_INTERVAL? What is the relation between DECAY_INTERVAL and CLOCK CYCLE?

    Sorry for taking you so many troubles. Hope for your response....


student

______________________________________

===================================================================
新浪免费电子邮箱 (http://mail.sina.com.cn)
新浪分类信息:轻松订阅,量身定制,好信息来找你! (http://classad.sina.com.cn/)
新浪闪烁短信闪亮登场 传情无限 (http://sms.sina.com.cn/cgi-bin/sms/smspic.cgi?spec=11&type=0)